Full-Stage Pseudo Label Quality Enhancement for Weakly-supervised Temporal Action Localization
This work addresses a key bottleneck in video analysis for researchers and practitioners by enhancing pseudo label quality, though it is incremental as it builds on existing frameworks.
The paper tackles the problem of improving pseudo label quality in weakly-supervised temporal action localization by proposing a framework with enhancement mechanisms at three stages, achieving an average mAP of 50.8% on THUMOS'14, which is a 1.2% improvement over the previous best method.
Weakly-supervised Temporal Action Localization (WSTAL) aims to localize actions in untrimmed videos using only video-level supervision. Latest WSTAL methods introduce pseudo label learning framework to bridge the gap between classification-based training and inferencing targets at localization, and achieve cutting-edge results. In these frameworks, a classification-based model is used to generate pseudo labels for a regression-based student model to learn from. However, the quality of pseudo labels in the framework, which is a key factor to the final result, is not carefully studied. In this paper, we propose a set of simple yet efficient pseudo label quality enhancement mechanisms to build our FuSTAL framework. FuSTAL enhances pseudo label quality at three stages: cross-video contrastive learning at proposal Generation-Stage, prior-based filtering at proposal Selection-Stage and EMA-based distillation at Training-Stage. These designs enhance pseudo label quality at different stages in the framework, and help produce more informative, less false and smoother action proposals. With the help of these comprehensive designs at all stages, FuSTAL achieves an average mAP of 50.8% on THUMOS'14, outperforming the previous best method by 1.2%, and becomes the first method to reach the milestone of 50%.