Séminaire de l’équipe MSDMA, vendredi 23 Février 2024

Map Unavailable

Date(s) - 23/02/2024
11 h 00 - 12 h 00

CNAM, accès 35, salle 35.1.53


Speaker: Yanwen Zhang (Complex Data Analysis Laboratory, Beihang University)

Title: Representation Learning based on Givens transformation and its applications


Orthonormality is the foundation of a number of matrix decomposition methods. For example, Singular Value Decomposition (SVD) implements the compression by factoring a matrix with orthonormal parts and is pervasively utilized in various fields. Orthonormality, however, inherently includes constraints that would induce redundant information,  and make the manipulation of orthonormal matrix difficult. An enhanced version of SVD, namely E-SVD, is accordingly established to losslessly and quickly release constraints and recover the orthonormal parts in SVD. E-SVD has a wide range of application field. For data compression, E-SVD will reduce 25% storage units as SVD reaches its limitation and fails to compress data. For blind watermarking, E-SVD theoretically guarantees the full retrievability of the watermark in the absence of an attack. For object detection, E-SVD give a new perspective of figuring out spatial variation in a matrix, leading to a wider usage of matrix factorization methods in the domain of unsupervised object detection.

Informations pratiques : On pourra participer au séminaire à distance en cliquant sur ce lien Teams (Meeting ID: 347 924 377 784 Passcode: pzpVfJ)