Multi-view pedestrian recognition using shared dictionary learning with group sparsity

Publication Type:
Conference Proceeding
Citation:
Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), 2011, 7064 LNCS (PART 3), pp. 629 - 638
Issue Date:
2011-11-28
Full metadata record
Files in This Item:
Filename Description Size
Thumbnail2011001831OK.pdf404.26 kB
Adobe PDF
Pedestrian tracking in multi-camera is an important task in intelligent visual surveillance system, but it suffers from the problem of large appearance variations of the same person under different cameras. Inspired by the success of existing view transformation model in multi-view gait recognition, we present a novel view transformation model based approach named shared dictionary learning with group sparsity to address the problem. It projects the pedestrian appearance feature descriptor in probe view into the gallery one before feature descriptors matching. In this case, L1,∞regularization over the latent embedding ensure the lower reconstruction error and more stable feature descriptors generation, comparing with the existing Singular Value Decomposition. Although the overall optimization function is not global convex, the Nesterovs optimal gradient scheme ensure the efficiency and reliability. Experiments on VIPeR dataset show that our approach reaches the state-of-the-art performance. © 2011 Springer-Verlag.
Please use this identifier to cite or link to this item: