Learning Equivariant Structured Output SVM Regressors

Journal article

Equivariance and invariance are often desired properties of a computer vision system. However, currently available strategies generally rely on virtual sampling, leaving open the question of how many samples are necessary, on the use of invariant feature representations, which can mistakenly discard information relevant to the vision task, or on the use of latent variable models, which result in non-convex training and expensive inference at test time. We propose here a generalization of stru...

Expand abstract

Authors: Vedaldi, A; Blaschko, M; Zisserman, A; IEEE

• Publication status: Published
• Journal: 2011 IEEE INTERNATIONAL CONFERENCE ON COMPUTER VISION (ICCV)
• Pages: 959-966
• Publication date: 2011-01-01
• DOI: 10.1109/ICCV.2011.6126339
• Language: English