Best Paper at CAIP 2021 for our work on material image classification

Sixiang XU, PhD student in our lab, has been awarded for the best paper at the International Conference on Computer Analysis of Images and Patterns (CAIP 2021) on September 30th. His paper entitled "Deep Fisher Score Representation via Sparse Coding" proposes an original solution for orderless pooling of deep (CNN) features for texture or material classification. The main idea consists in inserting a sparse encoder module in an end-to-end trainable network, in order to better fit the distribution of the high dimensional deep features. Sixiang will defend his PhD entitled "Feature selection, sparse coding and normalization for material image classification" before the end of 2021 in Saint-Etienne. He is supervised by Prof. Alain Trémeau and Damien Muselet.