/Action-DQN

Temporal Action Detection with Deep Q-Network. This is the code of paper "SAP: Self-adaptive action proporal with reinforcement learning"

MIT LicenseMIT

Watchers