no code implementations • ICCV 2023 • Nicolas Aziere, Sinisa Todorovic
Our key novelty is that we augment the original training videos in the deep feature space, not in the visual spatiotemporal domain as done by previous work.
no code implementations • CVPR 2019 • Nicolas Aziere, Sinisa Todorovic
Our first contribution is in specifying the N-pair loss in terms of manifold similarities between deep features.