Yair Weiss (School of Computer Science and Engineering, The Hebrew University of Jerusalem)
Tuesday, 31.12.2013, 11:30
Compressed sensing (CS) refers to a branch of applied mathematics which is based on the surprising result whereby signals that are exactly “k-sparse” (i.e. can be represented by at most k nonzero coefficients in some basis) can be exactly reconstructed using a small number of random measurements. Since natural images tend to be sparse in the wavelet basis, one of the motivating examples of CS has always been to reconstruct high resolution images from a small number of random m...
[ Full version ]
Monday, 30.12.2013, 13:00
Modern data analysis often faces the task of extracting characteristic features from sets of elements characterized with some measurement assay or procedure. In molecular biology, for example, an experiment may lead to measurement results pertaining to genes and then questions are asked about the properties of genes for which these were high or low. This is an example, of course, and the set of elements does not have to be genes. They can be genomic regions, proteins, structures, ...
[ Full version ]
Wednesday, 25.12.2013, 14:00
This work introduces a new family of redundancy schemes called replicated erasure codes (REC), which combine the storage-space efficiency of erasure codes and the repair-traffic efficiency of replications. A formal model for analyzing the storage and repair-raffic costs under availability and persistency constraints is also developed. It is shown that under parameters that characterize common P2P environments, REC generally achieves better results than each of the two methods sepa...
[ Full version ]
Yuval Rochman (Tel Aviv University/Marvell)
Wednesday, 25.12.2013, 13:00
We consider the problem of how to place and
efficiently utilize resources in network environments. The setting
consists of a regionally organized system which must satisfy
regionally varying demands for various resources. The operator
aims at placing resources in the regions as to minimize the cost
of providing the demands. Examples of systems falling under
this paradigm are 1) A peer supported Video on Demand service
where the problem is how to place various video movies,...
[ Full version ]
Avi Wigderson (Institute for Advanced Study, Princeton)
Wednesday, 25.12.2013, 12:30
I will survey what is known about the complexity of arithmetic circuits computing polynomials and rational functions with non-commuting variables, focusing on recent results and open problems. Strangely enough, some elementary questions in commutative algebra seem to hold the key both to new lower bounds and new algorithms.
The talk is mainly based on several papers with Pavel Hrubes and Amir Yehudayoff....
[ Full version ]
Lior Gorodisky (EE, Technion)
Tuesday, 24.12.2013, 11:30
Mitral Regurgitation (MR) is a cardiac disorder, in which there is dysfunction of the mitral valve, >resulting in backwards regurgitation of blood from the left ventricle into the left atrium. The gold standard to assess MR is by echocardiography, using ultrasound and the Doppler effect, which are more qualitative than quantitative. Using Echo-Doppler, the velocity of the blood can be measured only for the component that is parallel to the direction of the ultrasound beam, necessi...
[ Full version ]
Sunday, 22.12.2013, 11:30
A reliable density estimation is hard to obtain in problems of high-dimensional data, especially when the sample
used for estimation is small. As a result, various studies have tried to find approximate solutions to this problem
by reducing it to a less general, and hopesfully solvable, form. One prominent approach in this direction is
estimating the \emph{minimum-volume set (MV-set)} of a distribution at level $\alpha$ instead of its density
function. (An MV-set at level $\al...
[ Full version ]
Thursday, 19.12.2013, 08:30
You are invited to a special Symposium on "Aspects of Formal
Verification" on the occasion of Prof. Shmuel Katz' retirement. The symposium
will be held on
Thursday, December 19 2013, at Taub Building for Computer Science, in room 337
(3rd floor).
Participation is free but requires
pre-registraion.
You are all invited.
...
[ Full version ]
Sagi Snir (Haifa University)
Wednesday, 18.12.2013, 12:30
The reconstruction of evolutionary trees is a fundamental
task in Biology. The exponentially increasing amount of available genomic date over thousands of genes and species, gave rise to the task of combining all this data for the sake of constructing a phylogeny (evolutionary tree) depicting the evolution of life on earth along several billions of years.
Since accurate reconstruction is limited
to few dozens of species, the supertree approach, aims at accura...
[ Full version ]
Shiri Chechik (Microsoft Research, Silicon Valley)
Wednesday, 18.12.2013, 11:30
In recent years, the practical need for fast retrieval of shortest path queries has significantly increased, in part due to the developing GPS navigation systems and other route planning softwares.
Classical shortest paths algorithms such as Dijkstra's algorithm return shortest path distance in almost linear time in the number of nodes. However, continent-sized road networks require something much faster. It is possible to achieve sublinear-time queries if preprocessin...
[ Full version ]
Tuesday, 17.12.2013, 11:30
Understanding human attention have interested researchers for decades. The early works come from different fields of psychology and separate the cognitive process into
several steps. The different models of attention in static scenes have emerged and
evolved into dynamic saliency. Along with that, there are extensive cinematographic
theories on how the scene should be watched, or filmed. And again, there is a long
term research interest in view selection for static and dynamic...
[ Full version ]
Tuesday, 17.12.2013, 10:30
In this talk we present a novel approach for predicate abstraction of programs running on relaxed memory models. Our approach consists of two steps.
First, we reduce the problem of verifying a program P running on a memory model M
to the problem of verifying a program P_M that captures an abstraction of M as part of the program.
Second, we show how to discover new predicates that enable verification of P_M.
The core idea is to extrapolate from the predicates used to verify P und...
[ Full version ]
Tuesday, 17.12.2013, 08:30
Kogan Auditorium, EE Meyer Building
You are ivnited to the T2med Conference on Social-Mobile-Cloud Meets Medicine @ Technion,
organizerd by Deborah Estrin (CS, Cornell NYC Tech), Shie Mannor (EE, Technion), Uri
Rosenschein (Medical School, Technion)
The conference will be held on Tuesday, December 17th, 2013
The world around us has been transformed by the brave new world of Social (networks)
– Mobile (smart phones). The phenomenon of ubiquitous programmable, hand held
sensor pla...
[ Full version ]
Kohei Hatano (Tokyo Institute of Technology)
Monday, 16.12.2013, 11:30
EE Bloomfield Building 527
We consider online prediction problems of combinatorial concepts. Examples of such concepts include s-t paths, permutations, truth assignments, set covers, and so on. The goal of the online prediction algorithm is to compete with the best fixed combinatorial concept in hindsight. A generic approach to this problem is to design an online prediction algorithm using the corresponding offline (approximation) algorithm as an oracle. The current state-of-the art method, however, is not ...
[ Full version ]
Yoav Y. Schechner (EE, Technion)
Sunday, 15.12.2013, 13:00
This talk is about ;multi-view imaging via participating media, particularly the 3D volumetric scattering
atmosphere and the 3D wavy water-air interface. Camera multi-views are
either looking down from outer-space, or looking up from underwater or the
ground level. Multiple views on scales of tens or hundreds of kilometers effectively
create a huge lightfield camera-system. This provides constraints for
recovering complex sce...
[ Full version ]
Thursday, 12.12.2013, 11:30
Snake-in-the-box code is a Gray code which is capable to detect
a single error. Gray codes are important in the context of
rank modulation scheme which was suggested recently for
representing information in flash memories. For a Gray code
in this scheme the codewords are permutations, two consecutive
codewords are obtained by using the "push-to-the-top" operation,
and the distance measure is defined on permutations. In this talk
the Kendall's $\tau$-metric is used. We pres...
[ Full version ]
Ross Anderson (Computer Laboratory, University Of Cambridge)
Thursday, 12.12.2013, 11:00
Security protocols are the foundations on which the digital age is
built. SSL/TLS is the basis for online commerce, email privacy and
much else; EMV is taking over the world of card payments; and
lesser-known protocols such as SSH and DNSSEC protect the
infrastructure. Stable, reliable platforms are the basis on which
others can innovate; but what happens when the platforms themselves
fail? We have so far seen about a dozen failures of SSL/TLS, and had
to patch them in very...
[ Full version ]
Wednesday, 11.12.2013, 12:30
Online Learning deals with playing repeated games against an adversary with the aim of minimising a quantity known as regret, and has attracted much attention in recent years due to its applications in algorithm design and machine learning. The computational bottleneck in the application of online learning algorithms is the computation of a projection of a point onto a convex set, which for many problems of interest, such as those that arise in combinatorial optimization and matri...
[ Full version ]
Ross Anderson (Computer Laboratory, University Of Cambridge)
Tuesday, 10.12.2013, 14:30
There have been serious tussles over health privacy in the UK, the USA and elsewhere over the safety and privacy of health IT systems. Many countries are moving medical records away from family doctors, clinics
and hospitals to central cloud systems. This can cut costs and increase resilience: Hurricane Katrina wiped out many New Orleans patients' records, but those covered by the Veterans' Administration
could walk into any VA hospital and find their files waiting for them. But...
[ Full version ]
Pavel Kisilev (IBM Research Haifa)
Tuesday, 10.12.2013, 11:30
In this talk we present various research topics in medical imaging and related to them computer vision methods that are developed at IBM Haifa Research Lab. In the first part of the talk we describe various medical imaging problems and briefly hint on their solutions. The second part of the talk will be devoted to a newly developed general method for figure-ground segmentation. The method combines a bottom up approach of generating multiple highly plausible labelings of an image w...
[ Full version ]
Monday, 09.12.2013, 16:30
Acronyms are words formed from the initial letters of a phrase. For example, "CIA" is an acronym that usually means "Central Intelligence Agency," but in other contexts could mean "Culinary Institute of America." Understanding acronyms is important for many natural language processing applications, including search and machine translation.
While hand-crafted acronym dictionaries exist, they are limited and require frequent updates. We developed a new machine learning method to au...
[ Full version ]
Sunday, 08.12.2013, 13:00
Three dimensional user interfaces (3DUI) have great potential in computer aided geometric design (CAGD), where users work in a virtual 3D space and perform 3D operations. The mouse, commonly used for such tasks, provides only 2D input. In order to be used within the virtual space, the input is transformed to the virtual 3D space by reverse projection. However, this transformation is ambiguous, and cannot be used unless it is projected onto an object or constrained in some other w...
[ Full version ]
Chris Severs (Ph.D.) and Ryan Hennig (eBay)
Tuesday, 03.12.2013, 11:30
Over the past few years, Hadoop has been exploding in popularity. Ryan Hennig will discuss an overview of the MapReduce paradigm used to process data in Hadoop 1.0, and how it's shortcomings have motivated the new YARN architecture in Hadoop 2.0, which is better suited to scenarios like graph processing and machine learning. Chris Severs will discuss how functional programming is particularly well-suited to computation in the Big Data world, including the only four functions you...
[ Full version ]
Alex Kopaigorodski (EE, Technion)
Tuesday, 03.12.2013, 11:30
We present two methods for signal reconstruction from discrete data based on Prolate Spheroidal Wave Functions (PSWF). The first method is suitable for band-limited signals where the samples are available only in a finite interval of the signal. In such a case, an inherent reconstruction error is introduced by the standard reconstruction method. By using PSWF, better reconstruction results can be obtained based on the fast decay of the eigenvalues of the PSWF when the signal is re...
[ Full version ]
Thursday, 28.11.2013, 12:30
TinyLFU is a frequency based cache admission policy that boost effectiveness of caches. TinyLFU decides, based on the recent access history, whether it is worth admitting an accessed object into the cache at the expense of the eviction candidate. Realizing this concept is enabled through a novel approximate LFU structure, which maintains an approximate representation of the access frequency of recently accessed objects. ...
[ Full version ]
Wednesday, 27.11.2013, 12:30
In this talk we show several results regarding Boolean functions with small spectral norm (the spectral norm of $f$ is the sum of the absolute value of its Fourier coefficients.) Specifically, we prove the following results for Boolean functions on n variables with spectral norm $A$.
1. There is a subspace $V$ of co-dimension at most $A^2$ such that $f|_V$ is constant.
2. $f$ can be computed by a parity decision tree of size $2^{A^2}n^{2A}$. (a parity decision tree is a decisio...
[ Full version ]
Boris Oklander (Imperial College London)
Wednesday, 27.11.2013, 11:30
Admission control is a classical topic in communications, but cognition as a means to enhance the performance of users and the network is relatively new. Cognitive Users’ (CU) can achieve better performance based on smart access to resources such as power and spectrum. However, adaptation to dynamically changing conditions of communication channels is a challenge for CUs. Acting as a secondary user (SU), a CU cedes priority to ongoing transmissions of PUs and accesses a channel ...
[ Full version ]
Boaz Nadler (Weizmann Institute of Science)
Tuesday, 26.11.2013, 11:30
Phase retrieval - namely the recovery of a signal from its absolute Fourier
transform is a problem of fundamental importance in many fields.
In 1-D Phase retrieval is ill-posed, admitting multiple solutions.
In 2-D it often admits a unique solution, but finding it is challenging.
Current iterative algorithms are computationally intensive and in general have no
theoretical guarantees on the properties of their obtained solution.
In this talk I'll present a novel...
[ Full version ]
Muli Ben-Yehuda (CS, Technion)
Monday, 25.11.2013, 18:30
Nested virtualization---running multiple hypervisors in virtual machines---has come a long way in recent years. Since we first published KVM nested virtualization on Intel platforms ("The Turtles
project", 2010), nested virtualization has made important strides into the mainstream. These days, all hypervisors support it to some
degrees, and hardware support specifically for nested virtualization is starting to appear. The first part of this talk will provide an
overview of nest...
[ Full version ]
Christine Pranzas & Bastian Pranzas (ModuleWorks GmbH, Germany)
Sunday, 24.11.2013, 12:30
It is getting more and more popular to use 5-axis CNC-Milling machines to create parts out of metal or wood. 5-axis machines are necessary to place revolving tools at any position in space. During the presentation we want to give you a short overview about creating automated, efficient and smooth tool path algorithms for roughing and finishing (SWARF). The first presentation will give you an overview about 5-axis SWARF and 6-axis chain saw machining. Within the second talk you wi...
[ Full version ]
Marco Pistoia (IBM New York)
Thursday, 21.11.2013, 10:30
Security auditing of industry-scale software systems mandates automation. Static taint analysis enables exhaustive tracking of data flows for detection of leakage and integrity violations, such as cross-site scripting, SQL injection, and log forging. Research in this area has taken two directions: program slicing and type systems. Both of these approaches suffer from a high rate of false findings, which limits the usability of analysis tools based on these techniques. Attempts ...
[ Full version ]
Ilan Komargodski (Weizmann Institute of Science)
Wednesday, 20.11.2013, 12:30
Average-case lower bounds for De Morgan Formula Size.
We give a function $h:{0,1}^n \to {0,1}$ such that every deMorgan formula of size $n^{3-o(1)}/r^2$ agrees with $h$ on at most a fraction of $1/2+2^{-\Omega(r)}$ of the inputs.
Our technical contributions include a theorem that shows that the ``expected shrinkage'' result of H{\aa}stad actually holds with very high probability, where the restrictions are chosen from a certain distribution that takes into ac...
[ Full version ]
Alex Shpiner (EE, Technion)
Wednesday, 20.11.2013, 11:30
Data center networks form one of the most challenging network environments for routers, due to their fast link rates, short propagation times, and high performance demands. In this talk, we analyze two router functions related to packet processing: address resolution and order preservation.
The first part of the talk deals with VM (virtual machine) address resolution in data centers. Data centers can run multiple VMs, and potentially place them on any of the servers. The...
[ Full version ]
Marc Stevens (Centrum Wiskunde & Informatica, Netherlands)
Wednesday, 20.11.2013, 10:00
Flame, a highly advanced malware for cyberwarfare discovered in May, spread itself as a properly, but illegitimately, signed Microsoft update security patch. Flame achieved this by forging a signature from Microsoft using a so-called chosen-prefix collision attack on the very weak cryptographic hash function MD5. In this talk I will focus on counter-cryptanalysis, a new paradigm for strengthening cryptographic primitives, and the first example thereof, namely an efficient anomaly ...
[ Full version ]
Ronald Coifman (Mathematics and Computer Science, Yale)
Tuesday, 19.11.2013, 11:30
We will provide and overview of a range of ideas from Analysis, Statistical learning and diffusion Geometry , which address the transition from local data models to global configurations. In particular if the data under consideration is sensor data which is a noisy nonlinear complex transformation of natural parameters , we describe a methodology for deriving these parameters. Specifically we describe the role of eigenvectors (of appropriate locally learned “differential” tr...
[ Full version ]
Ari Trachtenberg (Boston University)
Sunday, 17.11.2013, 11:00
The ability to share and reconcile similar data on remote devices with little communication is fundamental to a wide variety of applications, ranging from maintenance of your contacts across smartphones to over-the-air programming of sensors and synchronous maintenance of files within a cloud. The problem also has interesting connections to information theory, cryptography, gene sequencing, and computational linguistics.
This two-talk course will begin with a formal state...
[ Full version ]
Gernot Heiser (University of New South Wales)
Thursday, 14.11.2013, 13:00
Henry Taub Distinguished Visitor Lecture: ...
[ Full version ]
Thursday, 14.11.2013, 12:30
abstract:Much good work fails to get the recognition it deserves because it does not get published in high-profile conferences. The good venues are highly competitive, and PC members have to review a large number of papers, and generally look for deficiencies to eliminate papers. For a paper to be accepted, it must not only have good ideas, it requires a convincing case that it addresses a relevant problem, and does a convincing job of solving the problem. Besides that, it mus...
[ Full version ]
Noa Avigdor-Elgrabli (CS, Technion)
Wednesday, 13.11.2013, 12:30
An Optimal Randomized Online Algorithm for Reordering Buffer Management
In the reordering buffer management problem an input
sequence of colored items arrives online, and has to be rescheduled
in a permuted output sequence of the same items, with the help
of a buffer that can hold $k$ items. The items enter the buffer
in their order of arrival. When the buffer is full, one item must be
removed and scheduled in the output sequence, making room for
a new input i...
[ Full version ]
Gernot Heiser (University of New South Wales)
Wednesday, 13.11.2013, 11:30
Henry Taub Distinguished Visitor Lecture: The L4 microkernel has undergone 20 years of use and evolution. It has an active user and developer community, and there are commercial versions which are deployed on a large scale and in safety-critical systems. In this paper we examine the lessons learnt in those 20 years about microkernel design and implementation. We revisit the L4 design papers, and examine the evolution of design and implementation from the original L4 to the latest ...
[ Full version ]
Guy Gilboa (EE, Technion)
Tuesday, 12.11.2013, 11:30
new framework is proposed for variational analysis and processing. It defines a
functional-based nonlinear transform and inverse-transform. The framework is
developed in the context of total-variation (TV), but it can be generalized to other
one-homogeneous functionals.
An eigenfunction, with respect to the subdifferential of the functional, such as a
disk in the TV case, yields an impulse in the transform domain. This can be viewed
as a generalization of know...
[ Full version ]
Gernot Heiser (University of New South Wales)
Monday, 11.11.2013, 18:30
Henry Taub Distinguished Visitor Lecture: A thorough and convincing evaluation is a core part of every systems paper, and generally involves some form of benchmarking. In my experience as a reviewer of hundreds of papers, there are a number of things people frequently get wrong in their evaluation. I have distilled those into a list of systems benchmarking crimes, most of which are sufficient reason to reject a paper. I will discuss those crimes and how to avoid them.
...
[ Full version ]
Gernot Heiser (University of New South Wales)
Monday, 11.11.2013, 12:30
Henry Taub Distinguished Visitor Lecture: With the formal verification of the seL4 microkernel, and subsequent work on assuring its safety and security properties, NICTA has recently not only demonstrated that unprecedented levels of assurance are possible, but also that the cost is competitive. In this talk I will examine what has been achieved, what the cost was, and how this might apply to larger systems, in particular the feasibility of assuring full-system safety or security....
[ Full version ]
Ari Trachtenberg (Boston University)
Sunday, 10.11.2013, 11:00
The ability to share and reconcile similar data on remote devices with little communication is fundamental to a wide variety of applications, ranging from maintenance of your contacts across smartphones to over-the-air programming of sensors and synchronous maintenance of files within a cloud. The problem also has interesting connections to information theory, cryptography, gene sequencing, and computational linguistics.
This two-talk course will begin with a formal state...
[ Full version ]
Merav Parter (Weizmann Institute of Science)
Wednesday, 06.11.2013, 12:30
We addresses the problem of designing a sparse {\em fault-tolerant} BFS tree, or {\em FT-BFS tree} for short, namely, a sparse subgraph $T$ of the given network $G$ such that subsequent to the failure of a single edge or vertex, the surviving part $T'$ of $T$ still contains a BFS spanning tree for (the surviving part of) $G$. Our main results are as follows. We present an algorithm that for every $n$-vertex graph $G$ and source node $s$ constructs a (single edge failure) FT-BFS tr...
[ Full version ]
Tuesday, 05.11.2013, 12:30
The CS Department and CeClub invite you to a special lecture by Jonathan "Peli" de Halleux (Senior Research SW Development Eng. Microsoft Research)
on
TouchDevelop: An App Creation Environment for a World of Devices, as follows:
We are experiencing a technology shift: powerful and easy-to-use mobile devices like smartphones and tablets are becoming more prevalent than traditional PCs and laptops. TouchDevelop is a novel programming environment and language...
[ Full version ]
Yong Joon Kim (CS, Technion)
Sunday, 03.11.2013, 13:00
We present an interactive-speed algorithm for computing the precise convex hull of freeform geometric models. The algorithm is
based on two pre-built data structures: (i) a Gauss map organized in a hierarchy of normal pyramids and (ii) a Coons bounding
volume hierarchy (CBVH) which effectively approximates freeform surfaces with a hierarchy of bilinear surfaces. For the axis
direction of each normal pyramid, we sample a point on the convex hull boundary using the CBVH. Th...
[ Full version ]
Thursday, 31.10.2013, 12:30
Advisor: Roy Friedman Abstract: This work presents MOLStream, a modular framework for rapid development
and evaluation of P2P live streaming systems. MOLStream allows P2P
streaming protocols to be decomposed into basic blocks, each associated with
a standard functional specification. By exposing structural commonalities between
these components, MOLStream enables specific implementations of these
building blocks to be combined in order to devise, refine and evaluate new P...
[ Full version ]
Wednesday, 30.10.2013, 13:00
In information science, ontology is a structured framework for organizing information which represents knowledge as a set of concepts and the relationships between them.
In this talk we will present our method which, given a domain, produces a multilingual ontology, that represents the main concepts within the domain, along with their attributes and the relationships between these concepts. Our process includes developing, relying on foundation ontologies, domain-ontology in Engl...
[ Full version ]
Uri Stemmer (Ben-Gurion University)
Wednesday, 30.10.2013, 12:30
Characterizing the Sample Complexity of Private Learners
In 2008, Kasiviswanathan et al. defined private learning as a combination of PAC learning and differential privacy. Informally, a private learner is applied to a collection of labeled individual information and outputs a hypothesis while preserving the privacy of each individual. Kasiviswanathan et al. gave a generic construction of private learners for (finite) concept classes, with sample complexity logarithmic in the s...
[ Full version ]
Ari Trachtenberg (Boston University)
Wednesday, 30.10.2013, 11:30
CS Taub Build. Auditorium 1
This talk will cover some of the prominent attack surfaces at all abstraction layers of modern smartphones, including those based on bypassing application signatures, USB takeover, GPS updates, commandeering the GSM subsystem, and filtering data from SSD memory. In the second part of the talk, we will present some of our own research into possible defenses against these or other attacks.
Bio:
Ari Trachtenberg is a Professor of Electrical and Computer Engine...
[ Full version ]
Nadav Har'El (Cloudius Systems)
Monday, 28.10.2013, 18:30
These days, most applications running on virtual machines in the cloud run on top of Linux. We all love Linux, but as an all-encompassing operating system for everything from phones to supercomputers, Linux was never really designed for virtual machines; It is big and complex, and it offers features (such as multi-user and multi-process) which are today made redundant by the hypervisor and slow it down. Linux's APIs are many times set in stone by decades of legacy code. All these ...
[ Full version ]
Sunday, 27.10.2013, 10:30
Two programs are said to be mutually terminating if they terminate on
exactly the same inputs. We suggest a proof rule for proving mutual
termination of a given pair of functions f,f' and the respective
subprograms that they call under a free context. Given a (possibly partial)
mapping between the functions of the two programs, the premise of the rule
requires proving that given the same arbitrary input in, f(in) and
f'(in) call functions mapped in the mapping with the same ...
[ Full version ]
Valerie King (University of Victoria)
Wednesday, 23.10.2013, 12:30
Byzantine agreement is a fundamental problem of distributed computing which involves coordination of processors when a constant fraction are controlled by a malicious adversary.
I'll present the first algorithm with polynomial expected time in the asynchronous model where the adversary may corrupt processors adaptively and has full knowledge of the state of all processors.
Ben-Or's 1983 expected exponential time algorithm reduced the problem to a series o...
[ Full version ]
Victor Luchangco (Oracle Labs)
Wednesday, 23.10.2013, 11:30
Transactional memory is a promising mechanism for synchronizing concurrent programs. There has been a lot of research in the past decade in implementing transactional memory. However, relatively little attention has been paid to precisely specifying what it means for them to be correct, which is necessary to reason rigorously about transactional memory.
I will discuss some of the issues with such specifications, and why there is not a single "correct" specification. ...
[ Full version ]
Ganesan Ramalingam (Microsoft Research India)
Tuesday, 22.10.2013, 11:30
Writing applications for distributed systems is challenging because of the pitfalls of distribution such as process failures and communication failures. One common requirement and challenge is the need for distributed applications that are idempotent. Idempotence ensures that the application functions correctly even when clients send duplicate requests, perhaps because the application failed to generate a response due to process failures, or because the response was generated but ...
[ Full version ]
Tomer Michaeli (Weizmann Institute of Science)
Tuesday, 22.10.2013, 11:30
Super resolution (SR) algorithms typically assume that the point spread function (PSF) of the camera is known, or else assumed to be a standard low-pass filter (e.g. a Gaussian). We demonstrate that the performance of such methods significantly deteriorates when the PSF deviates from their “one-size-fits-all” model. Deviations from the nominal PSF may be caused, e.g. by small camera shake inducing motion blur, by optical variations between sensing devices, etc. Such variations...
[ Full version ]
Sunday, 20.10.2013, 16:30
CS Taub Build. Auditorium 2
Dear students (girls, and also boys)
We invite you to launch a special project to promote women in Computer Science:
WOMEN AT CS.
The event will be held on Sunday, October 20, 2013 in Auditorium 2, CS Taub
Building for Computer Science.
Program:
16:30 - Greetings
- Prof. Irad Yavneh, Dean, Technion Computer Science
- Prof. Yossi Gil, Technion Computer Science
- Prof. Miriam Erez, Chair, Council for Advancing Women in Science a...
[ Full version ]
Sunday, 20.10.2013, 13:00
Applying a depth effect or a 3D-like look to hand-drawn animations can be just as visually
rewarding as rendering of a full 3D animation character. In this work we present a method
for computer-assisted shading of 2D animation sequences where the main character undergoes
non-rigid deformations between the frames.
The system receives as input a sequence of vectorized 2D animation frames containing an
outline sketch of the character and applies a 3D shading effect directly to ...
[ Full version ]
Wednesday, 16.10.2013, 14:30
In this talk we present practical sample complexity bounds for learning
Markov chains. In our setting a learning algorithm receives as input
the training data and the desired accuracy requirements and returns as
output a Markov chain that satisfies these requirements. An important
question is whether the training data contains enough information to achieve
the required accuracy. To answer this question we developed practical
lower bounds on the sample complexity of training ...
[ Full version ]
Tuesday, 15.10.2013, 11:30
Many signal and image processing applications have benefited remarkably from
the theory of sparse representations. In the classical synthesis model, the
signal is assumed to have a sparse representation under a given dictionary. In
this work we focus on greedy methods for the problem of recovering a signal
from a set of deteriorated linear measurements. We consider four different
sparsity frameworks that extend the aforementioned synthesis model that target
the signal's representa...
[ Full version ]
Orna Agmon Ben-Yehuda (CS, Technion)
Monday, 14.10.2013, 18:30
Physical memory is the scarcest resource in today’s cloud
computing platforms. Cloud providers would like to maximize
their clients’ satisfaction by renting precious physical
memory to those clients who value it the most. But real-world
cloud clients are selfish: they will only tell their providers
the truth about how much they value memory when it is in
their own best interest to do so. How can real-world cloud
providers allocate memory efficiently to those (selfish) ...
[ Full version ]
Yonathan Aflalo (EE, Technion)
Wednesday, 09.10.2013, 14:00
The field of shape analysis is rapidly growing. It involves processing of
geometric structures for which tools from numerical, metric, diffusion, and
differential geometries are exploited. In this thesis we propose to extend the set
of classical tools used in this domain by designing new geometries and solving
classical problems in the natural spectral domain. We first define a new scale-
invariant metric and extend existing procedures to deal with objects that are
subj...
[ Full version ]
Tuesday, 08.10.2013, 11:30
The goal of image oversegmentation is to divide an image into several pieces or "segments", such that each segment is part of an object present in the scene. Contrary to image segmentation algorithms, an oversegmentation algorithm is allowed to output more segments than the number of objects that appear in the image. Oversegmentation is a very common preprocessing step for several common computer vision tasks. In this work we study image oversegmentation and develop new algorithms...
[ Full version ]
Michael Baltaxe (Computer Science, Technion)
Tuesday, 08.10.2013, 11:30
The goal of image oversegmentation is to divide an image into several pieces or "segments", such that each segment is part of an object present in the scene. Contrary to image segmentation algorithms, an oversegmentation algorithm is allowed to output more segments than the number of objects that appear in the image. Oversegmentation is a very common preprocessing step for several common computer vision tasks. In this work we study image oversegmentation and develop new algorithms...
[ Full version ]
Steffen Weißmann (Technische Universität Berlin)
Monday, 07.10.2013, 13:00
Vortex filaments provide an intuitive and compact tool for the simulation of smoke in the context of special effects. We present a model that includes moving obstacles with vortex shedding as well as variational filament reconnection. The method is able to simulate smoke in high detail at interactive rates, even over long simulation periods.
Furthermore, we will present a method to decompose an existing velocity field (on a grid) into a set of vortex filaments, approximat...
[ Full version ]
Nir Hershko (Computer Science, Technion)
Sunday, 06.10.2013, 13:00
Maps of the world around us are used daily for many purposes such as navigation
or location-based services. However, our experience with maps is usually limited
to a 2D top-down projection. Using a 3D map instead have some advantages, such as
easier navigation and orientation, more realistic modeling and simulation of natural
phenomena such as water flow or line-of-sight analysis, and better understanding of
the elevation differences.
In this talk, we wil...
[ Full version ]
Wednesday, 02.10.2013, 11:30
Tables that match a given key to its value (e.g., hash tables) have become crucial algorithmic
building blocks for contemporary networking devices that need to take decisions on a large amount
of data at high speed. Unlike traditional table-based data structures, the networking environment
provides only very limited memory and necessitates a strict worst-case operation time.
Furthermore, since such tables typically lie on the critical path of networking devices, the tota...
[ Full version ]
Marc Donner (Engineering Director, Google)
Tuesday, 01.10.2013, 11:00
Strict implementation of data protection to support privacy protections for highly sensitive personal data can lead to difficult challenges for the development teams when faced with diagnostic, maintenance, and product evolution tasks. In this talk I will outline one specific privacy regime, healthcare data, that illustrates fundamental principles that will have to be addressed by all privacy-sensitive systems in the long term.
Bio:
Marc Donner was born in ...
[ Full version ]
Yonatan Zilpa (See Security College)
Monday, 30.09.2013, 19:30
In this talk we will give a brief review on SDN paradigm for networking.
We start with a brief historical introduction on the evolution of SDN.
While still an evolving SDN paradigm offers significant benefits. In
oppose to Interdomain routing the SDN paradigm separates the control
plane from the data plane this paradigm creates new opportunities and
challenges. Among the benefit side SDN offers better routing policy
constraints, easy service deployment, adaptability, securit...
[ Full version ]
Jonathan J. Klinger (Legal Counsel, Digital Rights Movement)
Monday, 16.09.2013, 18:30
In my lecture, we shall discuss Licensing basics, including (i) basis to
copyright law; (ii) what is software licensing; (iii) how does software
licensing differ from physical matters; (iv) what is open source
licensing. After grabbing the basics, we shall discuss the specialties
of Open Source software, and understand the different licenses and
license types.
...
[ Full version ]
Jonathan J. Klinger (Legal Counsel, Digital Rights Movement)
Monday, 02.09.2013, 18:30
In this talk, I shall discuss the use of biometrics for personal
identification by governmental authorities: The establishment of
biometric databases and the use of biometrics as unique identifiers
against state actors and other establishments. My main hypothesis is
that the use of biometrics is an apparatus for control that was first
used on inmates, criminals and other risk groups, then conveyed to
immigrants and other weak points in society. After this was concluded,
the...
[ Full version ]
Gur Harary (EE, Technion)
Wednesday, 14.08.2013, 11:30
My research addresses two fields. The first field is 3D spirals, were we
define two novel three-dimensional spirals: the 3D Euler spirals and the 3D
logarithmic spirals, and demonstrate their utility for curve completion and
for modeling. The second field is surface completion. Both fields are
incorporated to create a surface completion algorithm that is able to
complete extensively large holes.
In this talk I will focus on surface completion.
Physical a...
[ Full version ]
Raeda Naamneh (CS, Technion)
Sunday, 11.08.2013, 13:00
We propose an algorithm for reconstructing a multi-label object from cross-sections in a fair manner. We handle the problem in its full generality: Cross-sections need not be parallel or complete, every section may contain an unlimited number of contours with any geometries and in any level of containment hierarchy. We focus on the simultaneous reconstruction of an object from contours with multiple labels "colors", in scenarios in which interpolating separately between contours o...
[ Full version ]
Alexandra Gilinsky (EE, Technion)
Tuesday, 06.08.2013, 11:30
Computing distances between large sets of SIFT descriptors
is a basic step in numerous algorithms in computer vision.
When the number of descriptors is large, as is often the
case, computing these distances can be extremely time consuming.
In this research we propose the SIFTpack: a compact
way of storing SIFT descriptors, which enables significantly
faster calculations between sets of SIFTs than the current
solutions. SIFTpack can be used to represent SIFTs densely
extrac...
[ Full version ]
Monday, 05.08.2013, 18:30
Our new ability to create large amounts of biological data has created an abundance of tools to deal with the manipulations and data-mining required. however, like every new tool, they have limitations, edge-cases, and quirks. I've discovered that sometimes the old ways are the best ways. grep, sort, awk, and sed deal very will with large data-sets and parallel processing, and I've found that i use them for 90% of bioinformatic tasks, and quite successfully. I will describe the ty...
[ Full version ]
Micha Feigin (MIT Media Lab)
Monday, 05.08.2013, 11:30
The emergence of commercial time-of-flight (ToF) cameras has
motivated extensive utilization in computer vision. Gesture recognition,
scene identification, and depth estimation are important applications that
require accurate ranging information. The current prevailing approach in ToF
uses amplitude-modulated continuous wave illumination of a scene. These ToF
cameras produce real-time range maps at a relatively low cost. However, they
are geared to measure range (or phase) f...
[ Full version ]
Yohay Swirski (EE, Technion)
Tuesday, 23.07.2013, 11:30
The underwater environment is challenging for computer vision tasks. Poor
visibility, geometrical distortions and spatio-temporal varying illumination
are some of the physical sources of these challenges. Spatio-temporal
varying illumination is created by refraction of light through the wavy
water surface and is known as underwater flicker. In past studies, flicker
has often been considered to be an undesired effect, which degrades the
quality of images. In contrast, this re...
[ Full version ]
Monday, 22.07.2013, 18:30
This work addresses the scalability and efficiency of RAM-based storage systems wherein multiple objects must be retrieved per user request. Here, much of the CPU work is per server transaction, not per requested item. Adding servers and spreading the data across them also spreads any given set of requested items across more servers, thereby increasing the total number of server transactions per user request. The resulting poor scalability, dubbed the Multi-get Hole, has been repo...
[ Full version ]
Sunday, 21.07.2013, 09:15
Auditorium 2, CS Taub Building
The 2nd TCE
Summer Course on Computer Security will be held on Sunday-Thursday, July
21-25, 2013, at CS Taub Building, Auditorium 2.
Lectures will be given in Hebrew.
Attendees should hold a basic background in security in the for...
[ Full version ]
Raya Khanin (Bioinformatics Core, Memorial Sloan-Kettering Cancer Center, NY)
Wednesday, 17.07.2013, 13:30
As the production of next generation sequencing data is now a relatively
trivial component, it is apparent that bioinformatics is becoming the
single largest ongoing cost in terms of computing infrastructure and
personnel. Moreover, with dozens of computations methods being developed
for each step in the data processing, analysis and integration, it is
proving to be extremely challenging to design efficient and robust
pipeline for next generation sequencing data. This proble...
[ Full version ]
Rouven Strauss (CS, Technion)
Sunday, 14.07.2013, 13:00
We present a general, unified framework to resolve geometric covering problems. The problem is reduced to a set-cover search in parametric space. Both an optimal and an approximation
algorithm are proposed and implemented, each of which is highly parallel. Computer graphics techniques are employed as part of the framework's solution, and for higher efficiency, our implementation of this framework heavily exploits GPU based computations.
Our results are demonstrated b...
[ Full version ]
Katia Obraczka (UC Santa Cruz)
Wednesday, 10.07.2013, 12:30
This talk with provide an overview of the research on wireless self-organizing networking we have been conducting at UC Santa Cruz's Internetwork Research Group (i-NRG). I will describe some of our current projects ranging from mobility modeling and characterization, sensor network deployment, estimating network conditions using machine learning techniques.
Bio:
Katia Obraczka is Professor of Computer Engineering at UC Santa Cruz. Before joining UCSC, she held a r...
[ Full version ]
Hanan Samet (Department of Computer Science, University of Maryland)
Tuesday, 09.07.2013, 14:30
The recent introduction of the Apple iPhone 5 and the accompanying
iOS6 software environment which, among other changes, replaced the use of
a mapping App based on Google's map data with one that makes use of
Apple's map data, as well as changing the decisions as to what data is
displayed (served to the user) in responses to queries (especially
implicit ones through the manipulation of the viewing window),
has led to significant changes in the user experience with apps that
...
[ Full version ]
Yehuda Dar (EE, Technion)
Tuesday, 09.07.2013, 13:30
The digital video is a 3D spatio-temporal signal formed as a sequence of 2D
images (frames) captured over time. Hence, a sampled video represents a
large amount of information. As a result, video transmission and storage
systems require efficient coding and should be analyzed from a
rate-distortion perspective. Specifically, good quality video coding for low
bit-rate applications has great importance for transmission over
narrow-bandwidth channels and for storage with limite...
[ Full version ]
Tuesday, 09.07.2013, 11:30
One of the most important aspects of solving a problem is that of choosing an
appropriate parameterization. This trivial observation can be seen in many
forms in image processing and computer vision. Global parametrizations
include the Hough and Fourier transforms, whereas local parameterizations
include sparsity-based patch models and over-parameterized approaches. My
research explores important cases in motion analysis and 3D reconstruction
where a careful choice of the paramete...
[ Full version ]
Guy Rosman (CS, Technion)
Tuesday, 09.07.2013, 11:30
One of the most important aspects of solving a problem is that of choosing an
appropriate parameterization. This trivial observation can be seen in many
forms in image processing and computer vision. Global parametrizations
include the Hough and Fourier transforms, whereas local parameterizations
include sparsity-based patch models and over-parameterized approaches. My
research explores important cases in motion analysis and 3D reconstruction
where a careful choice of ...
[ Full version ]
Monday, 08.07.2013, 18:30
The Domain Name System (DNS) is key to the availability and correct operation of the Internet. Due to its significance it is also a lucrative target for attacks, most notably for cache poisoning. DNS cache-poisoning enables attackers to redirect clients to malicious hosts, allowing distribution of malware, credentials theft, phishing and spam, web sites defacement, and more.
Cryptographic defenses were designed (DNSSEC), but are not widely deployed; instead, multiple challenge-r...
[ Full version ]
Qiao WANG (Nanjing Institute of Communications Technologies)
Sunday, 07.07.2013, 14:30
There are over one hundred big cities in China, each with population more than one million. It brings us both challenge and opportunity to develop information processing technologies for establishing the urban service system, based on understanding the behavior of people in these cities. In this talk, we will illustrate the evolution patterns concerning the business hot-pots, traffic, and/or flu propagation, according to analysis from Chinese social networks and other sources. T...
[ Full version ]
Thursday, 04.07.2013, 12:30
Advisor: Erez Petrank Abstract: Efficient memory management of dynamic non-blocking data structures remains an
important open question. Existing methods either sacrifice the ability to
deallocate objects or reduce performance notably. In this paper, we present a
novel technique, called Drop the Anchor, which significantly reduces the
overhead associated with the memory management while reclaiming objects even in
the presence of thread failures. We demonstrate this memory ...
[ Full version ]
Moinuddin Qureshi (Georgia Institute of Technology)
Wednesday, 03.07.2013, 12:30
As conventional memory technologies such as DRAM run into the scaling wall, architects and system designers are forced to look at alternative technologies for building future computer systems. Several emerging Non-Volatile Memory (NVM) technologies such as PCM, STT-RAM, and Memristors have the potential to boost memory capacity in a scalable and power-efficient manner. However, these technologies are not drop-in replacements and will require novel solutions to enable their deploym...
[ Full version ]
Wednesday, 03.07.2013, 11:30
Spanner is Google's scalable, multi-version, globally-distributed, and synchronously-replicated database. It provides strong transactional semantics, consistent replication, and high performance reads and writes for a variety of Google's applications. I'll discuss the design and implementation of Spanner, as well as some of the lessons we have learned along the way. I'll also discuss some open challenges that we still see in building scalable distributed storage systems.
...
[ Full version ]
Tuesday, 02.07.2013, 10:30
The CS annual Program Engineering presentation will be held by the on Tuesday, July 2 2013, between 10:30-12.30, in CS
Taub 337 (3rd floor)as follows:
Elfy
Elfy is a software project designed to make the lives of cancer
patients and their loved ones better, by presenting:
An android application for monitoring various patients' everyday
health-related parameters combined with:
- real-time feedback to patients' ...
[ Full version ]
James Smith (University of Wisconsin, Emeritus)
Sunday, 30.06.2013, 14:30
A truly grand challenge for science in general, and for computer architects and designers in particular, is to understand the mammalian brain’s computing paradigm and then construct a computing device that embodies that paradigm. Although computer designers have a potential role to play in solving this grand challenge, it is up to us to define that role. From a computer designer’s perspective, I will illustrate the current understanding of the brain’s computational paradigm ...
[ Full version ]
Raanan Fattal (CS, Hebrew University of Jerusalem)
Sunday, 30.06.2013, 13:00
I will present a new multi-level preconditioning scheme for Laplacian matrices that arise in various computer vision and mathematical physics applications. The new approach combines principles from the multigrid and combinatorial preconditioning methodologies in order to overcome various shortcomings. In a first step, the new method adaptively eliminates a set of uncoupled variables from the linear system and by that simplifies the solution of nearly half of the variables in the s...
[ Full version ]
Friday, 28.06.2013, 09:00
Computer Architecture Day by TCE on Future Directions in Computer Architecture will be held on Friday, June 28 2013 in EE Meyer Building, Room 1003 between 9:00-15:00.
Leading researchers in Computer Architecture share their vision, outlook and views with students, faculty and industry representatives. This is A rare opportunity to hear from the best in the field.
...
[ Full version ]
Ronen Abravanel (Physics, Technion) and Daniel Vainsencher (EE, Technion)
Wednesday, 26.06.2013, 16:00
If you're like us, you love LyX, and wish you could do everything from it: - Write your papers with the conference style file - Collaborate with LaTeX-only people - Use it for a blog...
Problem: not everything works. Solution: report bugs and fix problems together, and eat pizza!
The plan on 26/6, 4 to 8 pm, in room 815 of the EE Mayer building:
- Lightning talks. Currently planned (you are invited to show us anything cool and not well known also!): - How to write Feynman di...
[ Full version ]
Wednesday, 26.06.2013, 13:30
With the expansion of cloud computing, more and more global services are
offered from several locations around the globe. Naturally, the demand for
cloud services in each geographical location changes over time depending
on the time of the day. Thus, when one data center (say in the east coast
of the US) experiences peak daily load, other data centers (say in Europe)
experience lower daily loads. Our research addresses the efficiency of load
sharing between geographically spread c...
[ Full version ]
Wednesday, 26.06.2013, 12:30
The 2013 Segel-Sudent End of Year Party will be held on Wednesday, June 26, 2013, between 12:30-14:30, at the CS Taub Lobby. ree beer, volleyball game with CS Team (on the lawn) plenty of fun are guaranteed. You are all invited. ...
[ Full version ]
Elisha Rosensweig (Infocom Best Paper Award)
Wednesday, 26.06.2013, 11:30
Over the past few years Content-Centric Networking, a networking model in which host-to-content communication protocols are introduced, has been gaining much attention.
A central component of such an architecture is a large-scale interconnected caching system. To date, the way these Cache Networks operate and perform is still poorly understood.
In this work, we demonstrate that certain cache networks are non-ergodic in that their steady-state characterization depends on the init...
[ Full version ]
Jennifer Sartor (Ghent University)
Monday, 24.06.2013, 11:30
Analyzing multi-threaded program scalability and performance on modern multicore machines is challenging, but paramount to continue optimizing software and to use hardware resources efficiently. Synchronization between threads results in some threads running while others wait on locks and barriers. We create a new metric, the criticality metric, to judge each thread's contribution to program execution time based on synchronization. The criticality metric takes into account how ...
[ Full version ]
Iddo Hanniel (Mechanical Engineering, Technion)
Sunday, 23.06.2013, 13:00
Solving polynomial systems of equations is an important problem in many fields such as computer-aided design, manufacturing and robotics. In recent years, subdivision-based solvers, which typically make use of the properties of the Bezier/B-spline representation, have proven very successful in solving such systems of polynomial constraints.
A major drawback in using subdivision solvers is their lack of scalability. When the given constraint is represented as a tensor p...
[ Full version ]
Henry Taub Distinguished Visitor Prof. Yale Patt (The University of Texas at Austin)
Friday, 21.06.2013, 09:00
Using computers to solve problems requires starting with a natural language formulation of the problem and systematically transforming it until one has a machine language (ISA) specification of the problem (i.e., a program). This then is executed on the implementation hardware. As Moore's Law continues to provide more and more transistors on a chip (50 billion transistors in a few years), application specialists continue to think up more and more applications that require addition...
[ Full version ]
Henry Taub Distinguished Visitor Prof. Yale Patt (The University of Texas at Austin)
Thursday, 20.06.2013, 14:30
Using computers to solve problems requires starting with a natural language formulation of the problem and systematically transforming it until one has a machine language (ISA) specification of the problem (i.e., a program). This then is executed on the implementation hardware. As Moore's Law continues to provide more and more transistors on a chip (50 billion transistors in a few years), application specialists continue to think up more and more applications that require addition...
[ Full version ]
Amir Herzberg (Bar-Ilan University)
Wednesday, 19.06.2013, 11:30
Everyone is concerned about Internet security, yet most traffic is not cryptographically protected. The usual justification is that most attackers are only off-path and cannot intercept traffic; hence, challenge-response mechanisms suffice to ensure authenticity. Usually, the challenges re-use existing `unpredictable' protocol header fields; this allows use of existing, widely-deployed protocols such as TCP and DNS.
We argue that this practice may only give an illusion ...
[ Full version ]
Daniel Zoran (Hebrew University of Jerusalem)
Tuesday, 18.06.2013, 11:30
Simple Gaussian Mixture Models (GMMs) learned from pixels of natural image
patches have been recently shown to be surprisingly strong performers in
modeling the statistics of natural images. Here we provide an in depth
analysis of this simple yet rich model. We show that such a GMM model is
able to compete with even the most successful models of natural images in
log likelihood scores, denoising performance and sample quality. We provide
an analysis of what such a model lear...
[ Full version ]
Monday, 17.06.2013, 18:30
Hypervisors implement useful features such as live migration and software-defined networking by interposing on their guest virtual machines’ I/O activity. Unfortunately, this interposition significantly reduces performance and scalability due to competition for resources between multiple guests and costly host/guest context switches. We present an efficient and scalable software-based I/O virtualization system that provides all of the benefits of I/O interposition while running ...
[ Full version ]
Danny Harari (Weizmann Institute of Science)
Monday, 17.06.2013, 11:30
We consider the tasks of learning to recognize hands and direction of gaze
from unlabeled natural video streams. These are known to be highly
challenging tasks for current computational methods. However, infants learn
to solve these visual problems early in development - during the first year
of life. This gap between computational difficulty and infant learning is
particularly striking. We present a model which is shown a stream of natural
videos, and learns without any sup...
[ Full version ]
Mira Shalah (CS, Technion)
Sunday, 16.06.2013, 13:00
Autodesk Maya provides powerful integrated animation, modeling, simulation, rendering, and compositing tools on a robust and extensible CG pipeline core. Autodesk’s Maya has proved invaluable to the CG industry. A skilled Maya modeler can build a computer model of a detailed machine part, or a multi-floor building, using just this flexible software. In fact, if you can imagine it, Maya can model it. The advanced Maya workflow extends further, taking the wireframe model and addin...
[ Full version ]
Henry Taub Distinguished Visitor Prof. Yale Patt (The University of Texas at Austin)
Friday, 14.06.2013, 09:00
Using computers to solve problems requires starting with a natural language formulation of the problem and systematically transforming it until one has a machine language (ISA) specification of the problem (i.e., a program). This then is executed on the implementation hardware. As Moore's Law continues to provide more and more transistors on a chip (50 billion transistors in a few years), application specialists continue to think up more and more applications that require addition...
[ Full version ]
Henry Taub Distinguished Visitor Prof. Yale Patt (The University of Texas at Austin)
Thursday, 13.06.2013, 14:30
Using computers to solve problems requires starting with a natural language formulation of the problem and systematically transforming it until one has a machine language (ISA) specification of the problem (i.e., a program). This then is executed on the implementation hardware. As Moore's Law continues to provide more and more transistors on a chip (50 billion transistors in a few years), application specialists continue to think up more and more applications that require addition...
[ Full version ]
Thursday, 13.06.2013, 12:45
Advisor:Roy FriedmanThis paper introduces a new family of redundancy schemes called replicated erasure codes (REC), which combine the storage-space efficiency of erasure codes and the repair-traffic efficiency of replications. A formal model for analyzing the storage and repair-traffic costs under availability and persistency constraints is also developed. It is shown that under parameters that characterize common P2P environments, REC generally achieves better results than ea...
[ Full version ]
Wednesday, 12.06.2013, 14:30
An error-correcting code with minimal distance d encodes a k-bit message by an n-bit codeword such that any two distinct codewords differ in at least d coordinates. It is well known that a random code, or even a random linear code, has a good minimal distance with high probability. Moreover, the conjectured intractability of decoding random linear codes has recently found several applications in cryptography.
A major disadvantage of random linear codes is that their encoding comp...
[ Full version ]
Wednesday, 12.06.2013, 13:00
In selective prediction, a predictor (e.g. classifier or regressor) is allowed to abstain from prediction on part of the domain. The objective is to improve the accuracy of predictions by compromising coverage. This talk will focus on the theoretical foundations of selective prediction and its applications for selective classification, selective regression, and active learning.
A classical result in statistical learning theory is that the excess risk of a classifier can be written...
[ Full version ]
Armando Castaneda (CS, Technion)
Wednesday, 12.06.2013, 12:30
The M-renaming task requires n+1 processes, each starting with a unique input name (from an arbitrary large range), to coordinate the choice of new output names from a range of size M. This talk presents the first upper bound on the complexity of 2n-renaming, when n + 1 is not a prime power.
It is known that 2n-renaming can be solved if and only if n+1 is not a prime power; however, the previous proof of the ``if" part was non-constructive, involving an approximation th...
[ Full version ]
Yale Patt (Henry Taub Distinguished Visitor)University of Texas at Austin
Wednesday, 12.06.2013, 11:30
Multicore has been around for several years now, and we hear it touted as the panacea of everything. ...until recently, that is. As expected, the hype has generated a lot of nonsense. We are told that multicore came about as a solution to a performance problem, that multicore allows you to run your problems at half the frequency and save power, that ILP is dead, that Moore's Law means we can put thousands (perhaps millions?) of cores on a single silicon die, that hardware works se...
[ Full version ]
Leonid Yaroslavsky (Tel-Aviv University)
Tuesday, 11.06.2013, 11:30
Transform image processing methods are methods that work in domains of image
transforms, such as Discrete Fourier, Discrete Cosine, Wavelet and alike.
They can be traced back to 1970-ths. The presentation reviews these methods
in their applications for image recovery and restoration, target location,
image re-sampling and interpolation. Specifically, the following issues are
addressed with emphasis on comparison and relationship with more recent and
popular methods:
- ...
[ Full version ]
Shai Avidan (Tel Aviv University)
Sunday, 09.06.2013, 13:00
This talk is more or less about depth. In the "more" part of the talk I will describe a method to recover depth and motion of a dynamic event captured by two snapshots of a camera array. The key idea is to represent the scene as a synthetic aperture volume and reduce the problem to volume registration that was already done in the medical imaging community. Camera arrays capture huge amounts of data and in the "less" part of the talk I will ask the question: how many images do we ...
[ Full version ]
Tobias Ritschel (Max Planck Institut informatik, Germany)
Sunday, 09.06.2013, 10:00
Creating visually convincing content, e.g. for movies or interactive applications such as computer games involves a lot of manual work, artistic skill and computational power. Ideally, however, users with limited artistic skills and competence, should be enabled to produce visually appealing content as well, in real-time and on every machine.
This talk will discuss four questions that are important when working towards this goal: What might be the fastest way to com...
[ Full version ]
Prof. Alexandru Losup (Delft University of Technology)
Thursday, 06.06.2013, 10:00
This series of talks focuses on the research and education conducted by the PDS group in recent years, on topics such as grid and cloud computing, systems for massively multiplayer online games, big data, and gamification of higher education.
Our research focuses on the modeling, the design, the implementation, and the analysis of parallel and distributed systems and algorithms. Our research is fundamental in that we aim to develop and evaluate generic methods and techni...
[ Full version ]
Wednesday, 05.06.2013, 15:30
Maintaining high availability of Ifrastructure-as-a-Service services at a reasonable cost is a challenging task that received recent attention due to the growing popularity of Cloud computing as a preferred means of affordable IT outsourcing. In large data-centers, faults are prone to happen and thus the only reasonable cost-effective method of providing high availability of services is an SLA aware recovery plan; that is, a mapping of the service VMs onto backup machines where th...
[ Full version ]
Michael Dinitz (Weizmann Institute of Science)
Wednesday, 05.06.2013, 12:30
In the matroid secretary problem we are given a stream of elements and asked to choose a set of elements that maximizes the total value of the set, subject to being an independent set of a matroid given in advance. The difficulty comes from the assumption that decisions are irrevocable: if we choose to accept an element when it is presented by the stream then we can never get rid of it, and if we choose not to accept it then we cannot later add it. Babaioff, Immorlica, and Klein...
[ Full version ]
Wednesday, 05.06.2013, 11:30
Handling noise and interference in wireless networks requires adaptive, high-performance error correction. Spinal codes are a new rateless error correcting code that iteratively applies a hash function to message bits, ensuring that two input messages that differ in even one bit produce very different coded sequences after the point at which they differ. Spinal codes offer a flexible tradeoff between computational cost and performance. Because spinal codes are rateless, they autom...
[ Full version ]
Prof. Alexandru Losup (Delft University of Technology)
Monday, 03.06.2013, 18:30
Hundreds of online games entertain over 250,000,000 online gamers in a maturing global market of over 30 billion Euros. A popular game genre is Real Time Strategy (RTS), which includes games such as StarCraft II (one of the best-selling games of 2010). Although Real-Time Strategy (RTS) games already entertain hundreds of millions of online players, the current games do not scale. For example, even popular RTS games such as the StarCraft series support in the same game instance onl...
[ Full version ]
Oliver Van Kaick (Tel Aviv University)
Sunday, 02.06.2013, 13:00
I would like to present our work where we address the challenge of computing consistent segmentations of sets of shapes. The sets may exhibit significant variability, with the shapes differing in their geometry and topology, as in the case of man-made shapes. We stipulate that, in this context, a correspondence between shape parts can be established by the incorporation of additional knowledge present in the set itself or provided by user input. Thus, we introduce three approaches...
[ Full version ]
Omri Azencot (CS, Technion)
Thursday, 30.05.2013, 13:00
Ph.D. Direct Track seminar:
We introduce a novel coordinate-free method for manipulating and analyzing
vector fields on discrete surfaces. Unlike the commonly used representations
of a vector field as an assignment of vectors to the faces of the mesh, or
as real values on edges, we argue that vector fields can also be naturally
viewed as operators whose domain and range are functions defined on the
mesh. Although this point of view is common in differential geo...
[ Full version ]
Thursday, 30.05.2013, 12:30
Advisor: Hagit Attiya Transactional memory (TM) has been hailed as a paradigm for simplifying concurrent programming.
While several consistency conditions have been suggested for TM, they are stated as conditions on possible histories that implementations may export. This leaves a striking disconnect with programming language abstractions providing atomic blocks to the programmer.
This paper presents the first connection between the observations exported by histories satisfyi...
[ Full version ]
Wednesday, 29.05.2013, 14:30
Epipolar geometry estimation is fundamental to many
computer vision algorithms. It has therefore attracted a lot of
interest in recent years, yielding high quality estimation
algorithms for wide baseline image pairs. Currently many types of
cameras (e.g., in smartphones and robot navigation systems) produce
geo-tagged images containing pose and internal calibration data.
Exploiting this information as part of an epipolar geometry estimation
algorithm may be useful but not trivial,...
[ Full version ]
Viki Mayzels (EE, Technion)
Wednesday, 29.05.2013, 11:30
Scale invariant feature detectors often find stable scales in only a few image pixels. Consequently, methods for feature matching typically choose one of two extreme options: matching a sparse set of scale invariant features, or dense matching using arbitrary scales. In this thesis we turn our attention to the overwhelming majority of pixels, those where stable scales are not found by standard techniques. We ask, is scale-selection necessary for these pixels, when dense, scale-inv...
[ Full version ]
Monday, 27.05.2013, 18:30
1. Resource management in Linux
2. Kernel Namespaces implementation (kernel 3.8)
3. Kernel Namespaces as an infrastructure for process virtualizaton
- Network namespaces and pid namespaces kernel implementation
- System calls for namespaces
- usage examples - creating linux namespaces in user space and attaching process
to these namespaces
4. cgroup kernel implementation
&...
[ Full version ]
Prof. Alexandru Losup (Delft University of Technology)
Monday, 27.05.2013, 10:00
This series of talks focuses on the research and education conducted by the PDS group in recent years, on topics such as grid and cloud computing, systems for massively multiplayer online games, big data, and gamification of higher education.
Our research focuses on the modeling, the design, the implementation, and the analysis of parallel and distributed systems and algorithms. Our research is fundamental in that we aim to develop and evaluate generic methods and techni...
[ Full version ]
Amir Vaxman (Vienna University of Technology, Austria)
Sunday, 26.05.2013, 12:00
Polyhedral meshes are meshes in which all faces are planar. These meshes are of paramount importance to architectural and industrial geometry, as such faces can be assumed with ease with plates of glass or wood. Furthermore, polyhedral meshes are considered as a discrete counterpart of continuous conjugate-parametrized surfaces. I will talk about methods to design such meshes by generalizing known methods from triangular-mesh processing, such as subdivision and deformation, and me...
[ Full version ]
Wednesday, 22.05.2013, 12:30
In large organizations, frequently the database has a large and intricate schema,
hence, formulating queries is cumbersome. In such organizations, users can benefit
from finding relevant queries in the query log of the database -- queries that
can serve as an initial example for query formulation, or queries that were written
by experts and elucidate how to write the query in an optimized manner. In this work
we describe a recommendation system that assists users by providing...
[ Full version ]
Shay Solomon (Weizmann Institute of Science)
Wednesday, 22.05.2013, 12:30
A {\em $(1+\eps)$-spanner} for a doubling metric $(X,\delta)$ is a subgraph $H$ of the complete graph corresponding to $(X,\delta)$, which preserves all pairwise distances to within a factor of $1+\eps$. A natural requirement from a spanner is to be robust against node failures, so that even when some of the nodes in the network fail, the remaining part would still provide a $(1+\eps)$-spanner. The spanner $H$ is called a {\em $k$-fault-tolerant $(1+\eps)$-spanner}, for any $0 \le...
[ Full version ]
Wednesday, 22.05.2013, 11:30
Storage cloud systems achieve economies of scale by serving multiple tenants from a shared pool of servers and disks. This leads to the commingling of data from different tenants on the same devices. Typically, a request is processed by an application running with sufficient privileges to access any tenant's data; this application authenticates the user and authorizes the request prior to carrying it out. Since the only protection is at the application level, a single vulnerabilit...
[ Full version ]
Anat Levin (Weizmann institute of Science)
Tuesday, 21.05.2013, 11:30
Recent attempts to fabricate surfaces with custom reflectance functions
boast impressive angular resolution, yet their spatial resolution is
limited. In this paper we present a method to construct spatially varying
reflectance at a high resolution of up to 220dpi, orders of magnitude
greater than previous attempts, albeit with a lower angular resolution. The
resolution of previous approaches is limited by the machining, but more
fundamentally, by the geometric optics model o...
[ Full version ]
Simon Korman (Tel-Aviv University)
Monday, 20.05.2013, 11:30
Fast-Match is a fast algorithm for approximate template matching under 2D
affine transformations that minimizes the Sum-of-Absolute-Differences (SAD)
error measure.
There is a huge number of transformations to consider but we prove that they
can be sampled using a density that depends on the smoothness of the image.
For each potential
transformation, we approximate the SAD error using a sublinear algorithm
that randomly examines only a small number of pixels...
[ Full version ]
Prof. Israel Koren (University of Massachusetts, Amherst)
Sunday, 19.05.2013, 16:00
The benefits of Asymmetric Multicore Processors (AMPs) vs. their symmetric counterparts highly depend on the thread scheduling followed. Since the computing needs of threads often vary during their execution, a fixed thread-to-core assignment is sub-optimal. Reassigning threads to cores (e.g., thread swapping) when a thread starts a new program phase (with different computational needs), can significantly improve the energy efficiency of AMPs.
Although identifying phase changes...
[ Full version ]
Sunday, 19.05.2013, 13:00
Glass and plywood are commonly used materials in architectural and industrial design. However,
these materials impose serious constraints on the creative freedom of a designer, as the space
of these constructions is the space of polyhedral meshes (PM), namely, meshes with planar
faces, which is known to be very limiting. In addition to the inherent difficulty in
generating such meshes, manipulating them while preserving planarity presents additional
challenges. Herein, we dis...
[ Full version ]
Monday, 13.05.2013, 18:30
This lecture will help to beginners to understand android architecture from the Linux perspective view. We will talk about the differences between Linux kernel and Android kernel, Android security model and Android initialization process....
[ Full version ]
Thursday, 09.05.2013, 13:30
The ocean covers 70% of the earth surface, and influences almost every aspect in our life, such as climate, fuel, security, and food. All over the world, including Israel, depleting resources on land are encouraging increased human activity in the ocean, for example: gas drilling, desalination plants, port constructions, aquaculture, bio-fuel, and more. The ocean is a complex, vast, foreign environment that is hard to explore and therefore much about it is still unknown. Interesti...
[ Full version ]
Thursday, 09.05.2013, 12:30
Advisor:Eran Yahav
Deterministic application replay has many use cases like reproducing bugs in long-running applications. A key challenge for deterministic replay is to record and replay asynchronous events like signals or context switches between threads in parallel programs. Unfortunately, existing approaches either do not support the deterministic replay of asynchronous events, are hard to deploy, or incur a record and/or replay overhead of 100x or more. rr is a determinis...
[ Full version ]
Xue-Cheng Tai (Mathematics, University of Bergen, Norway)
Thursday, 09.05.2013, 11:30
In this talk, we will explore the relationship between graph cut, convex
relaxation and some recent continuous max-flow approaches proposed in the
literature. Especially, we will see the continuous versus discrete
relationship between them. There are two advantages with these approaches:
1) different fast numerical algorithms have been used for these approaches;
2) global minimization can be guaranteed for some nonconvex problems. We
shall describe the connecti...
[ Full version ]
Prof. Alexandru Losup (Delft University of Technology)
Thursday, 09.05.2013, 10:00
This series of talks focuses on the research and education conducted by the PDS group in recent years, on topics such as grid and cloud computing, systems for massively multiplayer online games, big data, and gamification of higher education.
Our research focuses on the modeling, the design, the implementation, and the analysis of parallel and distributed systems and algorithms. Our research is fundamental in that we aim to develop and evaluate generic methods and techni...
[ Full version ]
Amir Yehudayoff (Technion)
Wednesday, 08.05.2013, 12:30
Communication complexity deals with the amount of
communication between parties needed to achieve a common goal. It was introduced by Yao and has found numerous applications since. We shall discuss a direct product theorem for 2-party communication. That is,
for example, if at least C bits of communication is needed to compute
a function f with probability at least 2/3, then communicating much
less than roughly C n^{1/2} bits to compute n copies of f yields
exponentially sm...
[ Full version ]
Wednesday, 08.05.2013, 12:30
In recent years, smartphones have become one of the most popular tools for local search and navigation.
Yet, state-of-the-art navigation tools are still rather limited in their ability to cope with complex route-planning tasks,
such as constructing an effective route from the location of the user to a target via a restaurant, a pharmacy and a coffee shop.
Planning becomes even more challenging when taking into account uncertainties,
such as the imprecise compliance of geograph...
[ Full version ]
Amit Berman (EE. Technion)
Wednesday, 08.05.2013, 11:30
The invention of semiconductor technology has marked a new era in memory devices: SRAM, DRAM, Flash and more. The ever-increasing rate of data production and consumption stimulates the development of high-performance memory devices. At times, high-density scaling drives new applications and ways of operation. However, device advancement presents trade-offs and ever growing challenges. High-performance memory (SRAM, DRAM) suffers from relatively low density, higher power consumptio...
[ Full version ]
Elisha Sacks (Purdue University)
Tuesday, 07.05.2013, 11:30
Computational geometry has developed efficient algorithms for a range of tasks
that could benefit computer graphics. Yet the practical application of these
algorithms is limited. One explanation is that the algorithms are correct and
efficient based on a theoretical analysis in the real-RAM model, whereas
graphics requires software that is reliable and efficient in practice. This so
called robustness problem has been attacked in several ways. I will describe
our solution...
[ Full version ]
Prof. Alexandru Losup (Delft University of Technology)
Tuesday, 07.05.2013, 10:00
This series of talks focuses on the research and education conducted by the PDS group in recent years, on topics such as grid and cloud computing, systems for massively multiplayer online games, big data, and gamification of higher education.
Our research focuses on the modeling, the design, the implementation, and the analysis of parallel and distributed systems and algorithms. Our research is fundamental in that we aim to develop and evaluate generic methods and techni...
[ Full version ]
Monday, 06.05.2013, 12:30
In the reordering buffer management problem an input sequence of colored items arrives online, and has to be rescheduled in a permuted output sequence of the same items, with the help of a buffer that can hold k items. The items enter the buffer in their order of arrival. When the buffer is full, one item must be removed and scheduled in the output sequence, making room for a new input item to enter the buffer. The objective is to minimize the total number of color changes between...
[ Full version ]
Prof. László Lovász (Eötvös Loránd University in Budapest, Hungary)
Thursday, 02.05.2013, 14:30
The limit theory of bounded-degree graphs is very interesting, but substantially more challenging than the dense theory. Limit objects can be defined in more than one sense; an interesting class of infinite graphs, which are called graphings and have been known from group theory and ergodic theory for a while, can be used to describe limit objects. Algorithmic questions in this theory are closely related to distributed computing in constant time.
The lecture is for mathe...
[ Full version ]
Thursday, 02.05.2013, 12:30
Advisor:Eran Yahav
We present an automated approach to concurrency control based on a restricted version of atomic sections that addresses a common and frequent need: the need to compose multiple atomic ADT operations into a new atomic operation.
Our approach aims to use and exploit the semantics of ADT operations to obtain specialized concurrency control solutions that can perform better than generic approaches. This is achieved by implementing ADT-specific locking, and ...
[ Full version ]
Prof. László Lovász (Eötvös Loránd University in Budapest, Hungary)
Wednesday, 01.05.2013, 13:30
the classical algorithmic problems must be rephrased, and a new kind of complexity theory emerges. Much of this has been worked out in the theory of "Graph Property Testing" in computer science, where we assume that information about such graphs is obtained by an appropriate sampling procedure. Graph limits offer a new perspective on this field, along with some new results.
Many questions in extremal graph theory can be phrased like this: what is the maximum of a certa...
[ Full version ]
Ariel Gabizon (CS, Technion) - CANCELLED
Wednesday, 01.05.2013, 12:30
Kuznetsov and Tsybakov considered the problem of storing information in a memory where a certain
p-fraction of the n cells are `stuck' at certain values. The person writing in the memory - the `encoder'-
knows which cells are stuck, and to what values. The person who will read the memory later - the `decoder'
is required to retrieve the message encoded without the information about which cells are stuck.
Kuznetsov and Tsybakov showed there are schemes where a message of length...
[ Full version ]
Michael Schapira (The Hebrew University of Jerusalem)
Wednesday, 01.05.2013, 11:30
New types of applications and services (from video streaming to cloud computing) are placing tremendous demands on today's networks. To tackle this challenge, network operators do traffic engineering (TE), i.e., tune routing-protocol parameters so as to use network resources efficiently. I will present recent (algorithmic and experimental) results regarding today's prevalent TE technique (ECMP routing) and also novel frameworks for TE (e.g., for TE with migration). I will discuss ...
[ Full version ]
Wednesday, 01.05.2013, 09:00
Command-line utilities is the real power of a Linux computer, and yet there's always a few useful utilities that even an experienced Linux user doesn't know about. This meeting is dedicated to going through several random utilities, some of which are less known, others are known but have surprising possibilities.
The range goes from things as simple as using "touch" to fake a modification date to using ssh for surfing the web (with your browser) as if your computer was in anothe...
[ Full version ]
Prof. László Lovász (Eötvös Loránd University in Budapest, Hungary)
Tuesday, 30.04.2013, 14:30
If you have a very large network (which may be deterministic or the result of some random procedure), we may want to approximate it by a smaller object, or by an infinite (analytic) object. The former question is related to Szemer\'edi's Regularity Lemma and its variants, the latter, to "graph limits". A theory of convergent graph sequences and their limits has been worked out by Benjamini and Schramm (for graphs with bounded degree) and by Borgs, Chayes, Lov\'asz, S\'os, Szegedy ...
[ Full version ]
Thierry Fraichard (INRIA Grenoble Rhône-Alpes)
Monday, 29.04.2013, 14:30
Autonomous navigation technologies have matured and improved to the
point that, in the past few years, self-driving cars have been able
to safely drive an impressive number of kilometers. It should be
noted though that, in all cases, the driver seat was never empty: a
human driver was behind the wheel, ready to take over whenever the
situation dictated it. This is an interesting paradox since the
point of a self-driving car is to remove the most unr...
[ Full version ]
Wednesday, 24.04.2013, 15:00
The defining characteristic of cloud computing platforms is money. In
clouds, non-cooperative clients pay their providers for the shared
computing resources they use as they use them. The introduction of
monetary compensation thus gives rise to a host of new possibilities
for efficiently sharing computing resources. We investigate the
economic foundations of cloud computing systems and propose new
mechanisms for non-cooperative clients and providers to share cloud
resources effici...
[ Full version ]
Liam Roditty (Bar-Ilan University)
Wednesday, 24.04.2013, 12:30
Computing distances is one of the most fundamental computational problems. In many applications we
are not really interested in all distances, we want the ability to retrieve them quickly. Thorup and Zwick
(2005) initiated the theoretical study of data structures capable of representing approximated distances
e?ciently, in terms of space requirement and query time.
Given an n-vertex weighted undirected graph with m edges, they show that for any integer k ≥ 1
it is...
[ Full version ]
Wednesday, 24.04.2013, 11:30
Real-time compression for primary storage is quickly becoming widespread as data continues to grow exponentially, but adding compression on the data path consumes scarce CPU and memory resources on the storage system. In this talk I'll present different approaches to efficient estimation of the potential compression ratio of data and how these methods can be applied in advanced storage systems. Our work targets two granularities: the macro scale estimation which is backed up by an...
[ Full version ]
Prof. Bernard Chazelle (Princeton University)
Tuesday, 23.04.2013, 14:30
The glory of 20th century physics was in many ways
the triumph of mathematics. Lacking the requisite symmetries,
the life sciences of today are unlikely to witness a repeat of this
miraculous match. Unlike electromagnetism, for example, cancer will
not be explained by a few differential equations.
The high descriptive complexity of biology seems to call for a new language
--- not a language of equations but of algorithms.
The challenge is to find it and then decipher it w...
[ Full version ]
Noga Levy (Tel-Aviv University)
Tuesday, 23.04.2013, 11:30
Face recognition in unconstrained videos requires specialized tools beyond
those developed for still images: the fact that the confounding factors
change state during the video sequence presents a unique challenge, but also
an opportunity to eliminate spurious similarities. Luckily, a major source
of confusion in visual similarity of faces is the 3D head orientation, for
which image analysis tools provide an accurate estimation.
The method we propose belongs to a f...
[ Full version ]
Prof. Bernard Chazelle (Princeton University)
Monday, 22.04.2013, 14:30
Imagine a group of interacting agents (eg, people, computers, birds, bacteria)
subject to the attracting influence of the agents with which they communicate.
Assume further that each agent is entitled to its own, distinct
algorithm for deciding whom to listen to when.
The communication graph may thus evolve endogenously
in arbitrarily complex ways. We show that such
an "influence system" is almost surely convergent if
the communication is bidirectional and asymptotically...
[ Full version ]
Myung-Soo Kim (Seoul National University)
Sunday, 21.04.2013, 13:00
We present real-time algorithms for various geometric operations (collision detection, minimum and Hausdorff distance computation,
surface-surface intersection, convex hull and distance field generation) on freeform geometric models including NURBS curves and surfaces.
These algorithms are based on bounding volume hierarchies (BVHs) for freeform shapes, which are considerably more compact than
conventional ones for mesh models. For static NURBS models, we often require 1000 t...
[ Full version ]
Thursday, 18.04.2013, 12:30
Advisor:Eran Yahav
Studying the evolution of software is an emerging topic, receiving much attention and focus. Every line of code written these days is more likely to belong to an existing piece of software: fixing, adding or removing functionality. Ensuring the correctness of code patches was long ago identified as crucial for efficiently producing quality software and the most prominent method for checking version correctness is regression testing - a cumbersome and time co...
[ Full version ]
Yonatan Goldhirsh (CS, Technion)
Wednesday, 17.04.2013, 12:30
Property testing studies algorithms that distinguish input in a property P from inputs far from it with
only a few queries. There are many natural cases where any property tester for a property P must use a lot
of queries, but it is possible to partition P into a few sub-properties P1; P2; :::; Pk, such that distinguishing
inputs in Pi from inputs far from P requires only a small amount of queries. This is the case for several well
studied properties such as concatenated palin...
[ Full version ]
Yossi Oren (Cryptography and Network Security Lab, Tel-Aviv University)
Wednesday, 17.04.2013, 11:30
Radio-Frequency Identification (RFID) tags based on the Electronic Product Code (EPC) standard have been aggressively introduced into the global supply chain. These tags were designed as an upgrade to the familiar 14-digit Universal Product Code (UPC) barcode. Due to the low cost and limited energy budget available to these tags, it was traditionally considered impractical to apply any sort of cryptographic protection to these tags.
In this talk I will show how, contrar...
[ Full version ]
Thursday, 11.04.2013, 12:30
Advisor: Prof. Mooly SagivSoftware applications are becoming increasingly harder to parallelize. Modern software systems written in imperative languages like Java and C# typically manipulate complex heap data structures, consist of multiple layers of abstraction, and have input- and deployment-specific behaviors that affect their available parallelism.
This creates a significant challenge for parallelizing compilers, whose target transformations are normally conditioned on t...
[ Full version ]
Wednesday, 10.04.2013, 15:00
The study of combinatorial problems with submodular objective functions has attracted much attention recently, and is motivated by the principle of economy of scale, prevalent in real world applications. Moreover, submodular functions are commonly used as utility functions in economics and algorithmic game theory. From a theoretical perspective, submodular functions and submodular optimization play a major role in combinatorics, graph theory and combinatorial optimization. In this...
[ Full version ]
Wednesday, 10.04.2013, 13:30
Physical memory is the most expensive resource in use in today's
cloud computing platforms. Cloud providers would like to maximize
their clients' satisfaction by renting precious physical memory to
those clients who value it the most. But real-world cloud clients
are selfish: they will only tell their providers the truth about how
much they value memory when it is in their own best interest to do
so. Under these conditions, how can providers find an efficient
memory allocation tha...
[ Full version ]
Jonathan Yaniv (CS, Technion)
Wednesday, 10.04.2013, 12:30
In the online preemptive scheduling problem, a computing system (e.g., a server) receives job processing requests throughout time, each request characterized by an arrival time, deadline, size and value. The goal of the scheduler is to maximize the total value of fully-completed jobs. Job preemption is allowed, i.e., jobs may be paused and resumed from the point at which they were preempted.
In its most general form, the problem admits a polylogarithmic lower bound on...
[ Full version ]
Udi Weinsberg (Technicolor Research, Palo Alto)
Wednesday, 10.04.2013, 11:30
Many online services, such as recommender systems, email, and social networks
collect user data, which is then used for both personalization and monetization.
Although the latter enables services to be free, users are realizing that these services come at a hidden cost of potentially exposing their private data. In this talk I will show that even the common 5-star item-rating recommender system leaks private demographic information. Then, I will discuss methods for helping use...
[ Full version ]
Noam Aigerman (Department of Computer Science and Applied Mathematics, Weizmann Institute)
Sunday, 07.04.2013, 13:00
We present an algorithm that, given a piecewise-linear mapping of a tetrahedral mesh, generates a similar map that is guaranteed to have bounded aspect-ratio distortion, and also to be locally injective. We further prove that if the boundary is mapped bijectively, the mapping is globally bijective.
We display several uses of our algorithm, namely for tetrahedral mesh deformations, parameterizations and mesh improvement.
...
[ Full version ]
Prof. Eldad Haber (Mathematics and Earth & Ocean Science, University of British Columbia)
Sunday, 07.04.2013, 11:30
Many applications such as enhanced oil recovery, in situ combustion and
contaminant transport are complex processes that are governed by flow in
porous media dynamics.
Predicting these flows is difficult since many of the parameters in the
equations are unknown.
In this talk we will present a method to integrate geophysical imaging and
multiphase flow to better predict the flow. We show that the problem is
mathematically similar to the problem of super resol...
[ Full version ]
Avishay Tal (Weizmann Institute of Science)
Wednesday, 03.04.2013, 12:30
For Boolean functions f : {0, 1} → {0, 1} and g : {0, 1}m → {0, 1}, the function composition of f and
g denoted by f ◦ g : {0, 1}nm → {0, 1} is the value of f on n inputs, each of them is the calculation of
g on a distinct set of m Boolean variables. Motivated by previous works that achieved some of the best
separations between complexity measures such as sensitivity, block-sensitivity, degree, certi?cate complexity
and decision tree complexity we show that most of the...
[ Full version ]
Gal Badishi (Cyvera Ltd., Cyber Defense Solutions)
Wednesday, 03.04.2013, 11:30
Recent attacks on high-value targets, demonstrate how state-of-the-art
defenses fail to protect against APTs (Advanced Persistent Threats).
These victims spare no expense and appropriately deploy cutting-edge
defenses, such as firewalls, intrusion detection and prevention systems,
and anti-virus scanners, as well as novel approaches for detecting
zero-day exploits - yet these are ineffective at thwarting determined
attackers.
In this talk we examine the unsatisfying state...
[ Full version ]
Gilad Asharov (Bar-Ilan University)
Wednesday, 20.03.2013, 12:30
Abstract: It is well known that it is impossible for two parties to toss a coin fairly (Cleve, STOC 1986). This result implies that it is impossible to securely compute with fairness any function that can be used to toss a coin fairly. In this paper, we focus on the class of deterministic Boolean functions with finite domain, and we ask for which functions in this class is it possible to information-theoretically toss an unbiased coin, given a protocol for securely computing the ...
[ Full version ]
Wednesday, 20.03.2013, 12:30
Pose and motion estimation of a calibrated camera is a common application in the photogrammetric world. In most cases this is solved by combining motion estimation with connections between the images taken by the camera and external geographical data (Orthophoto, Digital Terrain Model etc). While the motion estimation can be solved automatically, by Bundle Adjustment for example, the connections to the geographical data are generated manually. In this paper, we propose a novel app...
[ Full version ]
Inbar Raz (Security Architect at Check Point)
Tuesday, 19.03.2013, 12:30
Reverse engineering is the art of discovering the secrets of how things work. Be it a computer program, a mobile device or a washing machine, all devices have secrets that can be used for anything from improving or altering their behavior m(odding/tweaking,) through stealing their trade secrets and technology, to cracking their defenses and protections. We will talk about what it means, see how it’s done and learn to respect it !Y(eah, baby)!
Bio:
Inbar has been...
[ Full version ]
Vadim Indelman (Georgia Tech)
Tuesday, 19.03.2013, 11:30
This talk will focus on efficient methods for single- and multi-robot
localization and structure from motion (SfM) related problems such as
mobile vision, augmented reality and 3D reconstruction. High-rate
performance and high accuracy are a challenge, in particular when
operating in large scale environments, over long time periods and in
presence of loop closure observations. This challenge is further
enhanced in multi-robot configurations, where communication and
computat...
[ Full version ]
Jonathan Ben-Avraham (tk Open Systems)
Monday, 18.03.2013, 18:30
Beginning kernel programmers spend a large portion of their time looking for information regarding specific kernel versions, patches, the history of particular files and about submitters. It is easy to specify the information needed, but there is no well-known howto or cheat-sheet for finding it. This lecture presents a cheat-sheet for getting answers to the most kernel hacking common queries. The lecture will be in Hebrew. The lecture materials will be in English....
[ Full version ]
Thursday, 14.03.2013, 12:30
Based on a paper presented at SPAA 2012.
Advisor: Joseph (Seffi) Naor
Abstract:
We consider a market-based resource allocation model for batch jobs in cloud computing clusters. In our model, we incorporate the importance of the due date of a job by which it needs to be completed rather than the number of servers allocated to it at any given time.
Each batch job is characterized by the work volume of total computing units (e.g., CPU hours) along with a bound on maximum d...
[ Full version ]
Michael Kolomenkin (EE, Technion)
Tuesday, 12.03.2013, 11:30
Curves on objects can convey the inherent features of the shape. Edge
detection in images has been a fundamental problem in computer vision from
its early days.
Edge detection on surfaces, on the other hand, has received much less attention.
This talk defines a new class of view-independent curves on surfaces,
denoted relief edges. In a nutshell, these curves are the loci of the
"strongest" inflections on the surface.
We also propose a general framewor...
[ Full version ]
Guy Edri (See Secure Consulting)
Monday, 04.03.2013, 18:30
The talk includes a workshop, in which the speaker's AP will be
hacked by the audiance. To participate, bring a laptop with a flash disk on
which there is a Pent-test distribution:
http://www.backtrack-linux.org/.
Legal information about hacking
Introduction
Terminology (Access points, encryptions etc.) ...
[ Full version ]
Prof. Michael Kazhdan (Johns Hopkins University)
Monday, 04.03.2013, 11:00
In these lectures we will take an in-depth look at the Poisson equation, with a focus on its use in the graphics community.
We will start by looking at the way in which a number of common gradient-domain image-processing techniques result in a Poisson-like equation (including stitching, contrast enhancement, and low dynamic-range compression). We will discuss common discretizations of the linear systems and will look, in detail, at implementations of a multigrid solver ...
[ Full version ]
Wednesday, 27.02.2013, 14:30
Stability of a numerical algorithm is an important factor in its analysis, since an error exceeding the bound can lead to disastrous results: for example, the Patriot missile failure in 1991 occurred due to rounding errors, explosion of the Ariane 5 rocket in 1996 was a result of an overflow error, and sinking of the Sleipner A offshore platform was caused by a combination of a numerical error and a physical error. Introduction of concurrency to numerical algorithms results in a s...
[ Full version ]
Tuesday, 26.02.2013, 11:30
Activity recognition in video is a major problem in computer vision, that integrates two main challenges. The first, is defining a robust and informative set of features. The second, is constructing a model that builds on the set of features and provides a distinctive representation for each action. In this work we suggest a solution to both challenges. Our features are based on an extension of the GIST descriptor to space-time. We analyze the properties of space-time GIST in the ...
[ Full version ]
Prof. Mario Nemirovsky (Barcelona Supercomputer Center)
Tuesday, 26.02.2013, 11:00
This talk is to motivate an interactive discussion on a couple of new research proposals that presently I have been pushing (DCR and FOG). The markets of embedded real-time devices and datacenters/clouds are destined to converge. The explosive growth of the Internet of Things (IoT) will rapidly push the need to run time-critical applications in the cloud, not only for reducing costs but also for simplifying management and achieving increased availability and efficiency. However, f...
[ Full version ]
Yaniv Aknin and Yaniv Ben-Zaken
Monday, 18.02.2013, 18:30
We've come a long way since NCSA Mosaic and httpd. Web developers' arsenal today encompasses a very broad range of tools and techniques, but the developers also have to deal with rising user expectations and complex requirements. This talk offers a bit of historic perspective and a lot of real world modern solutions used for building the world's largest website. ...
[ Full version ]
Prof. Michael Kazhdan (Johns Hopkins University)
Monday, 18.02.2013, 11:00
In these lectures we will take an in-depth look at the Poisson equation, with a focus on its use in the graphics community.
We will start by looking at the way in which a number of common gradient-domain image-processing techniques result in a Poisson-like equation (including stitching, contrast enhancement, and low dynamic-range compression). We will discuss common discretizations of the linear systems and will look, in detail, at implementations of a multigrid solver ...
[ Full version ]
Zohar Levi (CS, Technion)
Sunday, 17.02.2013, 13:00
I will describe two researches on shape modeling.
In the first research we present a new framework for real-time shape deformation with local shape preservation and volume control. Given a 3D object, in any form, one would like to manipulate the object using convenient handles, so that the resulting shape is a natural variation of the given object. It is also important that the deformation is controlled, thereby enabling localized changes that do not influence nearby branches.
...
[ Full version ]
Wednesday, 13.02.2013, 12:30
Over the last several years, modern biology has undergone an information revolution, which is apparent in a shift of thinking and practice. While typical biological studies were, until recently, mostly focused on specific pathways, like the p53 signaling pathway, the emergence of novel high-throughput technologies now enables the quantification of biological features in a genome-wide scale and on multiple levels of molecular regulation.
In this talk I will describe some of the me...
[ Full version ]
Ittay Eyal (EE, Technion)
Wednesday, 13.02.2013, 11:30
Advances in data center technologies have led to extensive use of data centers to store large volumes of data in a managed distributed system. The users of such systems have increasing expectations of both efficiency and reliability.
Recent years have shown that even the largest cloud storage providers occasionally fail, and users have to replicate data among multiple providers to obtain reliability. However, classical replication techniques (e.g., ABD) are not applica...
[ Full version ]
Octavia I. Camps (EE and CE, Northeastern University)
Tuesday, 12.02.2013, 16:30
Library Classroom, Aerospace Engineering
Cameras are ubiquitous everywhere and hold the promise of significantly changing the way we live and interact with our environment. Dynamic vision systems are uniquely positioned to address the needs of a growing segment of the population. Smart environments that are aware of user activities would enable an aging population to carry on independent lives
for as long as possible. The power of geometric invariants to provide solutions to computer vision problems towards realizing th...
[ Full version ]
Monday, 11.02.2013, 18:30
The "flash-friendly file system (F2FS)" is a new file system for Linux recently announced by engineers from Samsung. Unlike jffs2 and logfs, f2fs is not targeted at raw flash devices, but rather at the specific hardware that is commonly available to consumers ג SSDs, eMMC, SD cards, and other flash storage with a flash translation layer (FTL) already built in.
Required prior Knowledge will be covered by the preliminary lecture.
...
[ Full version ]
Dima Sezganov (EE, Technion)
Tuesday, 05.02.2013, 11:30
The Bag-Of-Features (BOF) approaches are becoming central in
large-scale image retrieval. The standard BOF method is orderless,
completely omitting geometric configuration of visual words. The
geometrical information is usually involved only in the
post-processing spatial verification step usually implemented with the
RANdom SAmple Consensus (RANSAC) algorithm. To enable visual search in
real-time, RANSAC can be applied only to a relatively small number of
top...
[ Full version ]
Monday, 04.02.2013, 18:30
Bitcoin is the world's first decentralized digital currency. Powered by a peer-to-peer network of computers running free software, it was created as an alternative to the existing monetary system in an attempt to offer people more autonomy over their finances and ease of transaction.
In this talk, we will give a broad overview of the Bitcoin system and describe its motivation and its current status. We will then explain the open cryptographic protocol in its foundation and how it...
[ Full version ]
Prof. Michael Kazhdan (Johns Hopkins University)
Monday, 04.02.2013, 11:00
In these lectures we will take an in-depth look at the Poisson equation, with a focus on its use in the graphics community.
We will start by looking at the way in which a number of common gradient-domain image-processing techniques result in a Poisson-like equation (including stitching, contrast enhancement, and low dynamic-range compression). We will discuss common discretizations of the linear systems and will look, in detail, at implementations of a multigrid solver ...
[ Full version ]
Monday, 04.02.2013, 10:30
Automata over infinite words are a core concept in program verification.
Many different types of automata exist and they can be classified by the type of their transition relation and by the type of their acceptance condition.
In this thesis we explore weak automata over infinite words.
In weak automata the state space is partitioned into partially ordered sets.
The transition relation is then restricted so that a state from one set can only move to states in partitions lower in t...
[ Full version ]
Tal Moran (Interdisciplinary Center Herzliya)
Wednesday, 30.01.2013, 12:30
Time-lock puzzles are a class of computational problems that take
longer to solve than they do to generate. They can be used, for
example, to send a message ``to the future'': the sender publishes a
puzzle whose solution is the message to be sent, thus hiding it for
the time it takes to solve the puzzle. Since adversaries may have
access to many more computers than honest solvers, massively parallel
solvers should not be able to produce a solution much faster than
serial on...
[ Full version ]
Vadim Ratner (EE, Technion)
Tuesday, 29.01.2013, 14:30
Diffusion-type algorithms have been integrated successfully into the toolbox used in image processing and computer vision. We introduce in the context of digital signal and image processing a new more flexible and powerful family of
parabolic-hyperbolic partial differential equations (PDEs) that model elastic deformation. The analogy between a well-known physical system and process on one hand, and the dynamics of an image processing scheme on the other hand, contributes inter...
[ Full version ]
Tuesday, 29.01.2013, 13:30
Although event processing is considered an emerging technology,
empirical studies show that most enterprises ability to use such systems as ranging from poor to fair.
Moreover, many of them indicated the skill issue as a barrier to adoption.
One of the major contributors to this observation is the inherent difficulty in enforcing temporal correctness using current languages and tools.
This difficulty can lead to incorrect outcomes or to work being hacked around these issues for a ...
[ Full version ]
Prof. Maks Ovsjanikov (École Polytechnique)
Sunday, 27.01.2013, 13:00
This talk discuss a representation of maps between pairs of 3D shapes (represented as triangle meshes) that generalizes the standard notion of a map to include correspondences that are not necessarily point-to-point. This representation is compact, and yet allows for efficient inference (shape matching) and enables a number of applications, including algebraic map manipulation such as computing map sums and differences.
The key aspect of this representation is that many...
[ Full version ]
Noa Avigdor-Elgrabli (CS, Technion)
Wednesday, 23.01.2013, 12:30
In the reordering buffer management problem (RBM) a sequence
of $n$ colored items enters a buffer with limited capacity $k$.
When the buffer is full, one item is removed to the output sequence,
making room for the next input item. This step is repeated until
the input sequence is exhausted and the buffer is empty. The
objective is to find a sequence of removals that minimizes the
total number of color changes in the output sequence. The
problem formalizes numerous applicati...
[ Full version ]
Armando Castaneda (CS, Technion)
Wednesday, 23.01.2013, 11:30
In a distributed task, processes of a distributed system start with private input values, communicate each other using a medium and then irrevocably decide an output value. In the M-renaming task each process of the system starts with a distinct input name. and must decide a distinct output name in the range [1, ..., M]. The renaming task is at the core of the theory of distributed computing.
Several papers claimed that, in an asynchronous distributed system in which p...
[ Full version ]
Monday, 21.01.2013, 18:30
Command-line utilities is the real power of a Linux computer, and yet there's always a few useful utilities that even an experienced Linux user doesn't know about. This meeting is dedicated to going through several random utilities, some of which are less known, others are known but have surprising possibilities.
The range goes from things as simple as using "touch" to fake a modification date to using ssh for surfing the web (with your browser) as if your computer was ...
[ Full version ]
Oren Freifeld (Applied Mathematics, Brown University)
Sunday, 20.01.2013, 14:30
Three-dimensional object shape is commonly represented in terms of deformations
of a triangular mesh from an exemplar shape. In particular, statistical
generative models of human shape deformation are widely used in computer
vision, graphics, ergonomics, and anthropometry. Existing statistical models,
however, are based on a Euclidean representation of shape deformations. In
contrast, we argue that shape has a manifold structure: For example, averaging
the shape deformations...
[ Full version ]
Renjie Chen (Computer Science, Technion)
Sunday, 20.01.2013, 13:00
The Delaunay triangulation and its dual, the Voronoi diagram, are among the most fundamental structures in computational geometry. We show how to localize the Delaunay triangulation of a given planar point set, namely, bound the set of points which are possible Delaunay neighbors of a given point. We then exploit this observation in an algorithm for constructing the Delaunay triangulation (and its dual Voronoi diagram) by computing the Delaunay neighbors (and Voronoi cell) of each...
[ Full version ]
Jens Groth (University College London)
Wednesday, 16.01.2013, 12:30
Abstract: Zero-knowledge proofs enable a prover to convince a verifier of the truth of a statement without revealing anything else. Non-interactive zero-knowledge proof are zero-knowledge proofs where the prover constructs the proof without interaction with the verifier.
In a probabilistically checkable proofs the verifier only needs to read a small number of bits to be convinced that the statement to be proven is true with high probability. PCPs have been used to reduc...
[ Full version ]
Dorit Ron (Weizmann Institute of Science)
Wednesday, 16.01.2013, 11:30
The Bitcoin scheme is a rare example of a large scale global payment system in which all the transactions are publicly accessible (but in an anonymous way). We downloaded the full history of this scheme, and analyzed many statistical properties of its associated transaction graph. In this paper we answer for the first time a variety of interesting questions about the typical behavior of users, how they acquire and how they spend their bitcoins, the balance of bitcoins they keep in...
[ Full version ]
Eli Appleboim (EE, Technion)
Tuesday, 15.01.2013, 11:30
A discrete version of the Ricci flow, applicable to images, is introduced and applied for some image processing tasks such as denoising and non-linear interpolation. This flow is unique among the geometric flows that have been applied in image processing, in the sense that it is the only flow wherein the metric of an image evolves, rather than the image itself, as is the case in other geometric flows. Our curvature and flow represent an adaptation of the combinatorial Ricci curvat...
[ Full version ]
Ola Svensson (EPFL, Switzerland)
Wednesday, 09.01.2013, 12:30
K-median is the problem where we wish to open k facilities so as to minimize the average distance each client has to its closest opened facility. The lack of progress on this central problem compared to facility location (a close relative) is partly due to the difficulty of handling the hard constraint that at most k facilities are allowed to be opened.
In this talk we shall see that we can relax this constraint into a soft constraint with a "violation-dependent" inc...
[ Full version ]
Rami Cohen (IBM R&D Lab, Haifa)
Wednesday, 09.01.2013, 11:30
Virtualizing resources for easy pooling and accounting, as well as for rapid provisioning and release, is essential for the effective management of modern data centers. Although the compute and storage resources can be virtualized quite effectively, a comprehensive solution for network virtualization has yet to be developed. Our analysis of the requirements for a comprehensive network virtualization solution identified two complimentary steps of ultimate importance. One is specify...
[ Full version ]
Yaron Kalit (EE, Technion)
Tuesday, 08.01.2013, 11:30
In recent years, a number of super-resolution techniques have been proposed.
Most of these techniques construct a high resolution image by either combining
several low resolution images at sub-pixel misalignments or by learning
correspondences between low and high resolution image pairs. These techniques
and most other image interpolation methods focus on grayscale images.
In this work, a new super-resolution method for color textures from a single
image is presented. The in...
[ Full version ]
Thursday, 03.01.2013, 12:30
We present a communication-efficient protocol for distributed online prediction tasks where a group of learners, such as computers in a cloud or nodes in a sensor network, provide a real-time service based on local models of a high-frequency data stream (as in, e.g., online advertisements, trade recommendations, or social content promotion). While synchronizing these local models can increase the predictive performance, the required communication generates a cost. In contrast to e...
[ Full version ]
Pavel Hrubes (University of Calgary)
Wednesday, 02.01.2013, 12:30
The main open problem in proof complexity is to prove superpolynomial lower bounds for the so-called Frege system, where Frege system is a specific formalization of classical propositional calculus. The same problem can be posed for non-classical logics, and I will discuss two popular examples: intuitionistic and modal logic. Here, one actually can prove exponential lower bounds on lengths of proofs. I will show how to construct this lower bound and discuss connections between pro...
[ Full version ]
Rotem Osman (University of Toronto)
Wednesday, 02.01.2013, 11:30
To date, all concurrent search structures that can support predecessor queries have had depth logarithmic in m, the number of elements. This work introduces the SkipTrie, a new concurrent search structure supporting predecessor queries in amortized expected O(log log u + c) steps, insertions and deletions in O( c log log u ), and using O(m) space, where u is the size of the key space and c is the maximum point contention during the operation.
The SkipTrie is a probabil...
[ Full version ]
Alon Faktor (The Weizmann Institute of Science)
Tuesday, 01.01.2013, 11:30
We define a "good image cluster" as one in which images can be easily composed (like a puzzle) using pieces from each other, while are difficult to compose from images outside the cluster. The larger and more statistically significant the pieces are, the stronger the affinity between the images. This gives rise to unsupervised discovery of very challenging image categories. We further show how multiple images can be composed from each other simultaneously and efficiently using a c...
[ Full version ]
