AI Paper Atlas

Research map / Vision-language and multimodal / Cross-modal learning

How Do Optical Flow and Textual Prompts Collaborate to Assist in Audio-Visual Semantic Segmentation?

Yujian Lee, Peng Gao, Yongqi Xu, Wentao Fan

ICCV 2025 · Vision-language and multimodal › Cross-modal learning · approach: Audio-visual source localization · 2 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
add optical-flow pre-mask (ablation)AAVS · AVSBench MS3mIoU: 67.3 -> 69.5Improved
add optical-flow pre-mask (ablation)AAVS · AVSBench S4mIoU: 83.2 -> 84.1Improved
add optical-flow pre-mask (ablation)AAVS · AVSBench AVSSmIoU: 48.5 -> 49.2Improved
add post-mask loss on top of pre-mask (ablation)AAVS + pre-mask · AVSBench MS3mIoU: 69.5 -> 70.2Improved
add post-mask loss to full model (ablation)SSP · AVSBench MS3mIoU: 70.4 -> 72.3Improved
add textual prompts without VTA (ablation)AAVS · AVSBench MS3mIoU: 67.3 -> 68.1Improved
add textual prompts without VTA (ablation)AAVS · AVSBench AVSSF-score: 53.2 -> 53.0No clear effect
add VTA to textual prompts (ablation)AAVS + textual prompts · AVSBench MS3mIoU: 68.1 -> 69.8Improved
add VTA to textual prompts (ablation)AAVS + textual prompts · AVSBench MS3F-score: 78.9 -> 83.3Improved
add textual prompts and VTA to pre-mask and post-mask (ablation)SSP · AVSBench MS3mIoU: 70.2 -> 72.3Improved
replace simple cross-attention with VTA (ablation)SSP · AVSBenchmIoU: +1.1% mIoU, +0.5% F-scoreImproved
improve textual prompt quality (worst A to best D) (ablation)SSP with VTA · AVSBench MS3mIoU: 2.3% and 1.2% in mIoU and F-scoreImproved

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy