Research map / Vision-language and multimodal / Cross-modal learning
Ref-AVS: Refer and Segment Objects in Audio-Visual Scenes
Yaoting Wang, Peiwen Sun, Dongzhan Zhou, Guangyao Li, Honggang Zhang, Di Hu
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 changed | Where | Measured | Effect |
|---|---|---|---|
| add text expression information (ablation) | EEMC · Ref-AVS Bench (Seen) | J (%): 23.67 -> 34.20 | Improved |
| add audio information (ablation) | EEMC · Ref-AVS Bench (Seen) | J (%): 27.27 -> 34.20 | Improved |
| add cached memory for temporal dynamics (ablation) | EEMC · Ref-AVS Bench (Unseen) | J (%): 36.73 -> 49.54 | Improved |
| add cached memory (ablation) | EEMC · Ref-AVS Bench (Unseen) | F-score: 0.660 -> 0.648 | Hurt |
| add modality encoding tokens [aud] and [vis] (ablation) | EEMC · Ref-AVS Bench (Mix) | J (%): 36.09 -> 41.87 | Improved |
| add audio to text-guided referring segmentation methods (ReferFormer +audio) | ReferFormer · Ref-AVS Bench (Seen) | J (%): 27.46 -> 31.31 | Improved |
| add text to AVS methods (AVSegFormer +text) | AVSegFormer · Ref-AVS Bench (Seen) | J (%): 29.16 -> 33.47 | Improved |
| use RoBERTa+VGGish instead of Flan-T5 or LanguageBind encoders (ablation) | EEMC · Ref-AVS Bench | mIoU: 35.17 -> 40.87 | Improved |
| increase expression difficulty grade (ablation) | EEMC · Ref-AVS Bench (by grade) | mIoU: 39.27 -> 33.39 | Hurt |
| add temporal cues to spatial expressions (ablation) | EEMC · Ref-AVS Bench (dimension analysis) | mIoU: 33.83 -> 41.01 | Improved |
Closest papers by meaning
- Think Before You Segment: An Object-aware Reasoning Agent for Referring Audio-Visual Segmentation (AAAI 2026 · similarity 0.91)
- TSAM: Temporal SAM Augmented with Multimodal Prompts for Referring Audio-Visual Segmentation (CVPR 2025 · similarity 0.91)
- SAM2-LOVE: Segment Anything Model 2 in Language-aided Audio-Visual Scenes (CVPR 2025 · similarity 0.90)
- Unraveling Instance Associations: A Closer Look for Audio-Visual Segmentation (CVPR 2024 · similarity 0.89)
- AURORA: Augmented Understanding via Structured Reasoning and Reinforcement Learning for Reference Audio-Visual Segmentation (AAAI 2026 · similarity 0.89)
- Improving Audio-Visual Segmentation with Bidirectional Generation (AAAI 2024 · similarity 0.88)
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.
