Generating action proposals in untrimmed videos is a challenging task, since video sequences usually contain lots of irrelevant contents and the duration of an action instance is arbitrary. The quality of action proposals is key to action detection performance. The previous methods mainly rely on sliding windows or anchor boxes to cover all ground-truth actions, but this is infeasible and computa- tionally inefficient. To this end, this paper proposes Recap- Net - a novel framework for generating action proposal, by mimicking the human cognitive process of understanding video content. Specifically, this RecapNet includes a resid- ual causal convolution module to build a short memory of the past events, based on which the joint probability ac- tionness density ranking mechanism is designed to retrieve the action proposals. The RecapNet can handle videos with arbitrary length and more importantly, a video sequence will need to be processed only in one single pass in order to generate all action proposals. The experiments show that, the proposed RecapNet outperforms the state-of-the- art under all metrics on the benchmark THUMOS14 and ActivityNet-1.3 datasets.
|Number of pages||11|
|Journal||IEEE Transactions on Cybernetics|
|Publication status||Accepted/In press - 2 Jan 2020|