אירועים והרצאות בפקולטה למדעי המחשב ע"ש הנרי ומרילין טאוב

Among the many ways we have to model signals, one approach that has found great popularity in the past decade is based on sparse representations. The main focus in our work is exploring novel sparsity-based signal models that go beyond the classical one. In this talk I will present two such contributions.

1) Statistical models based on sparse representations: In this work we introduce statistical dependencies between various components of the sparsity-based model. These include intra-dependencies within a single sparsity pattern and inter-dependencies between a pair of related representations. We have applied these models on natural image patches, shedding light on the statistics of this family of signals and utilizing them for developing a highly efficient scheme for single image super-resolution.

2) The analysis viewpoint for sparse representations: This is an alternative and different viewpoint to sparse representations, which has only recently started to attract attention. We have explored the dictionaries associated with this type of model, in terms of the dictionary properties leading to a desired recovery performance and the ability to learn such a dictionary from a set of signal examples.

PhD research under the supervision of Prof. Michael Elad (CS Department).