End-to-end Learning of Action Detection from Frame Glimpses in Videos
This work addresses the problem of efficient and accurate action detection in videos for computer vision applications, representing a novel method for a known bottleneck.
The paper tackles action detection in videos by introducing an end-to-end recurrent neural network agent that learns to predict temporal bounds of actions by observing and refining glimpses of frames, achieving state-of-the-art results on THUMOS'14 and ActivityNet datasets while observing only 2% or less of video frames.
In this work we introduce a fully end-to-end approach for action detection in videos that learns to directly predict the temporal bounds of actions. Our intuition is that the process of detecting actions is naturally one of observation and refinement: observing moments in video, and refining hypotheses about when an action is occurring. Based on this insight, we formulate our model as a recurrent neural network-based agent that interacts with a video over time. The agent observes video frames and decides both where to look next and when to emit a prediction. Since backpropagation is not adequate in this non-differentiable setting, we use REINFORCE to learn the agent's decision policy. Our model achieves state-of-the-art results on the THUMOS'14 and ActivityNet datasets while observing only a fraction (2% or less) of the video frames.