CLASP: Cross-modal Salient Anchor-based Semantic Propagation for Weakly-supervised Dense Audio-Visual Event Localization

The Dense Audio-Visual Event Localization (DAVEL) task aims to temporally localize events in untrimmed videos that occur simultaneously in both the audio and visual modalities. This paper explores DAVEL under a new and more challenging weakly-supervised setting (W-DAVEL task), where only video-level event labels are provided and the temporal boundaries of each event are unknown. We address W-DAVEL by exploiting $\mathit{\text{cross-modal}}\mathit{\text{salient}}\mathit{\text{anchors}}$, which are defined as reliable timestamps that are well predicted under weak supervision and exhibit highly consistent event semantics across audio and visual modalities. Specifically, we propose a $\mathit{\text{Mutual}}\mathit{\text{Event}}\mathit{\text{Agreement}}\mathit{\text{Evaluation}}$ module, which generates an agreement score by measuring the discrepancy between the predicted audio and visual event classes. Then, the agreement score is utilized in a $\mathit{\text{Cross-modal}}\mathit{\text{Salient}}\mathit{\text{Anchor}}\mathit{\text{Identification}}$ module, which identifies the audio and visual anchor features through global-video and local temporal window identification mechanisms. The anchor features after multimodal integration are fed into an $\mathit{\text{Anchor-based}}\mathit{\text{Temporal}}\mathit{\text{Propagation}}$ module to enhance event semantic encoding in the original temporal audio and visual features, facilitating better temporal localization under weak supervision. We establish benchmarks for W-DAVEL on both the UnAV-100 and ActivityNet1.3 datasets. Extensive experiments demonstrate that our method achieves state-of-the-art performance.