Adaptive Mutual Supervision for Weakly-Supervised Temporal Action Localization

Weakly-supervised temporal action localization aims to localize actions from untrimmed long videos with only video-level category labels. Most previous methods ignore the incompleteness issue of Class Activation Sequences (CAS), suffering trivial detection results. To tackle this issue, we propose a novel Adaptive Mutual Supervision (AMS) framework two branches, where base branch detects most discriminative regions, while supplementary localizes less regions through an adaptive sampler. The sampler dynamically updates inputs for using sampling weight sequence negatively correlated CAS branch, thus encouraging underestimated by branch. promote mutual enhancement between further construct location supervision. Each adopts pseudo-labels generated other as By alternately optimizing branches multiple iterations, progressively complete regions. Extensive experiments on THUMOS14 and ActivityNet1.2 demonstrate that proposed AMS method significantly outperforms state-of-the-art methods.