Research map / Vision-language and multimodal / Cross-modal learning
Implicit Counterfactual Learning for Audio-Visual Segmentation
Mingfeng Zha, Tianyu Li, Guoqing Wang, Peng Wang, Yangyang Wu, Yang Yang, Heng Tao Shen
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 all three components (MIT, SC, CDCL) to the baseline (ablation) | ICF · AVS-Object M3 | J&F: 64.51 -> 69.89 | Improved |
| add all three components (MIT, SC, CDCL) to the baseline (ablation) | ICF · AVSS | J&F: 41.86 -> 48.16 | Improved |
| add semantic counterfactual (SC) on top of MIT and CDCL (ablation) | ICF · AVS-Object M3 | J&F: 67.63 -> 69.89 | Improved |
| add CDCL on top of MIT and SC (ablation) | ICF · AVSS | J&F: 45.66 -> 48.16 | Improved |
| add video and segment granularity to frame-level implicit text (ablation) | ICF · AVSS | J&F: 42.34 -> 44.10 | Improved |
| generate counterfactuals with diffusion in continuous space instead of discrete VQVAE (ablation) | ICF · AVS-Object M3 | J&F: 68.51 -> 69.89 | Improved |
| add L_ortho constraint to continuous counterfactual generation (ablation) | ICF · AVSS | J&F: 47.31 -> 48.16 | Improved |
| use inter-sample and intra-sample counterfactuals together instead of feature-level (ablation) | ICF · AVS-Object M3 | J&F: 67.63 -> 69.89 | Improved |
| use distribution-based contrast instead of prototype contrast (ablation) | ICF · AVSS | J&F: 46.22 -> 48.16 | Improved |
| replace implicit orthogonal counterfactuals with explicit Text Revision (ablation) | ICF · AVS-Object M3 | J&F: 69.89 -> 68.44 | Hurt |
| integrate ICF into AVSBench | AVSBench · AVS-Object M3 | J&F: 59.25 -> 62.30 | Improved |
| integrate ICF into AVSegFormer | AVSegFormer · AVSS | J&F: 39.33 -> 43.44 | Improved |
Closest papers by meaning
- Unraveling Instance Associations: A Closer Look for Audio-Visual Segmentation (CVPR 2024 · similarity 0.89)
- Unsupervised Audio-Visual Segmentation with Modality Alignment (AAAI 2025 · similarity 0.87)
- Looking Similar Sounding Different: Leveraging Counterfactual Cross-Modal Pairs for Audiovisual Representation Learning (CVPR 2024 · similarity 0.87)
- Robust Audio-Visual Segmentation via Audio-Guided Visual Convergent Alignment (CVPR 2025 · similarity 0.87)
- CF-VLM:CounterFactual Vision-Language Fine-tuning (NeurIPS 2025 · similarity 0.87)
- Dynamic Derivation and Elimination: Audio Visual Segmentation with Enhanced Audio Semantics (CVPR 2025 · similarity 0.87)
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.
