ActAR: Actor-Driven Pose Embeddings for Video Action Recognition.

Soufiane Lamghari, Guillaume-Alexandre Bilodeau, Nicolas Saunier

Browse the full CVPR paper archive.