HR-Pro: Point-supervised Temporal Action Localization via Hierarchical Reliability Propagation

TL;DR

HR-Pro introduces a hierarchical reliability propagation framework for point-supervised temporal action localization, leveraging high-confidence cues at snippet and instance levels to improve accuracy and outperform existing methods.

Contribution

The paper proposes a novel hierarchical reliability propagation framework that effectively utilizes point annotations at multiple levels for improved temporal action localization.

Findings

Achieves 60.3% mAP on THUMOS14 benchmark.

Outperforms previous point-supervised methods.

Surpasses several fully supervised approaches.

Abstract

Point-supervised Temporal Action Localization (PSTAL) is an emerging research direction for label-efficient learning. However, current methods mainly focus on optimizing the network either at the snippet-level or the instance-level, neglecting the inherent reliability of point annotations at both levels. In this paper, we propose a Hierarchical Reliability Propagation (HR-Pro) framework, which consists of two reliability-aware stages: Snippet-level Discrimination Learning and Instance-level Completeness Learning, both stages explore the efficient propagation of high-confidence cues in point annotations. For snippet-level learning, we introduce an online-updated memory to store reliable snippet prototypes for each class. We then employ a Reliability-aware Attention Block to capture both intra-video and inter-video dependencies of snippets, resulting in more discriminative and robust snippet representation. For instance-level learning, we propose a point-based proposal generation approach as a means of connecting snippets and instances, which produces high-confidence proposals for further optimization at the instance level. Through multi-level reliability-aware learning, we obtain more reliable confidence scores and more accurate temporal boundaries of predicted proposals. Our HR-Pro achieves state-of-the-art performance on multiple challenging benchmarks, including an impressive average mAP of 60.3% on THUMOS14. Notably, our HR-Pro largely surpasses all previous point-supervised methods, and even outperforms several competitive fully supervised methods. Code will be available at https://github.com/pipixin321/HR-Pro.