EAR: Enhancing Uni-Modal Representations for Weakly Supervised Audio-Visual Video Parsing

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Li, Huilai, Di, Xiaomeng, Xing, Ying, Dang, Yonghao, Wang, Yiming, Yin, Jianqin
Format: Preprint
Published: 2026
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866917476539826176
author Li, Huilai
Di, Xiaomeng
Xing, Ying
Dang, Yonghao
Wang, Yiming
Yin, Jianqin
author_facet Li, Huilai
Di, Xiaomeng
Xing, Ying
Dang, Yonghao
Wang, Yiming
Yin, Jianqin
contents Weakly supervised Audio-Visual Video Parsing (AVVP) aims to recognize and temporally localize audio, visual, and audio-visual events in videos using only coarse-grained labels. Faced with the challenging task settings, existing research advances along two main paths: pre-training pseudo-label generators for fine-grained cross-modal semantic guidance, or refining AVVP model architectures to enhance audio-visual fusion. However, since audio and visual signals are typically unaligned, achieving accurate video parsing fundamentally relies on precise perception of uni-modal events. Yet these multi-modal focused strategies excessively emphasize multi-modal fusion while inadequately guiding and preserving uni-modal semantics, resulting in noisy pseudo-labels and sub-optimal video parsing performance. This paper proposes a novel framework that enhances uni-modal representations for both the pseudo-label generator and the AVVP model. Specifically, we introduce a similarity-based label migration approach to annotate pre-training data, thereby enabling the pseudo-label generator to better understand uni-modal events. We also employ a soft-constrained manner to refine modeling of uni-modal features in parallel with multi-modal fusion. These designs enable coordinated attention to both uni-modal and cross-modal representations, thus boosting the localization performance for events. Extensive experiments show that our method outperforms state-of-the-art methods in both pseudo-label and AVVP performance.
format Preprint
id arxiv_https___arxiv_org_abs_2605_08723
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle EAR: Enhancing Uni-Modal Representations for Weakly Supervised Audio-Visual Video Parsing
Li, Huilai
Di, Xiaomeng
Xing, Ying
Dang, Yonghao
Wang, Yiming
Yin, Jianqin
Computer Vision and Pattern Recognition
Multimedia
Weakly supervised Audio-Visual Video Parsing (AVVP) aims to recognize and temporally localize audio, visual, and audio-visual events in videos using only coarse-grained labels. Faced with the challenging task settings, existing research advances along two main paths: pre-training pseudo-label generators for fine-grained cross-modal semantic guidance, or refining AVVP model architectures to enhance audio-visual fusion. However, since audio and visual signals are typically unaligned, achieving accurate video parsing fundamentally relies on precise perception of uni-modal events. Yet these multi-modal focused strategies excessively emphasize multi-modal fusion while inadequately guiding and preserving uni-modal semantics, resulting in noisy pseudo-labels and sub-optimal video parsing performance. This paper proposes a novel framework that enhances uni-modal representations for both the pseudo-label generator and the AVVP model. Specifically, we introduce a similarity-based label migration approach to annotate pre-training data, thereby enabling the pseudo-label generator to better understand uni-modal events. We also employ a soft-constrained manner to refine modeling of uni-modal features in parallel with multi-modal fusion. These designs enable coordinated attention to both uni-modal and cross-modal representations, thus boosting the localization performance for events. Extensive experiments show that our method outperforms state-of-the-art methods in both pseudo-label and AVVP performance.
title EAR: Enhancing Uni-Modal Representations for Weakly Supervised Audio-Visual Video Parsing
topic Computer Vision and Pattern Recognition
Multimedia
url https://arxiv.org/abs/2605.08723