Research map / Vision-language and multimodal / Cross-modal learning
Inconsistency-aware Multimodal Schrodinger Bridge for Deepfake Localization
Jiayu Xiong, Jing Wang, Qi Zhang, Wanlong Wang, Jun Xue
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 the witness bridge (WSB) for cross-modal filtering and budget allocation (ablation) | IaMSB · LAV-DF | AP@0.95: 28.16 -> 55.92 | Improved |
| add the refinement bridge (RSB) (ablation) | IaMSB · LAV-DF | AP@0.95: 23.39 -> 55.92 | Improved |
| add the coarse bridge (CSB) (ablation) | IaMSB · LAV-DF | AP@0.5: 83.25 -> 99.33 | Improved |
| raise coarse steps from 1 to 2 (ablation) | IaMSB CSB · LAV-DF | AP@0.95: 52.87 -> 55.92 | Improved |
| raise coarse steps from 2 to 3 (ablation) | IaMSB CSB · LAV-DF | AP@0.95: 55.92 -> 55.98 | No clear effect |
| use witness top-k of 16 rather than 2 (ablation) | IaMSB WSB · LAV-DF | AP@0.95: 50.98 -> 55.92 | Improved |
| widen witness top-k from 16 to 64 (ablation) | IaMSB WSB · LAV-DF | AP@0.95: 55.92 -> 53.97 | Hurt |
| IaMSB versus RegQAV on strict IoU | IaMSB · AV-Deepfake1M | AP@0.95: 12.57 -> 23.01 | Improved |
| IaMSB versus MMMS-BA on recall at AR@20 | IaMSB · LAV-DF | AR@20: 95.93 -> 95.41 | Hurt |
| IaMSB versus UMMAFormer on visual-only forgery strict IoU | IaMSB · TVIL | AP@0.95: 62.43 -> 65.62 | Improved |
Closest papers by meaning
- FreqDebias: Towards Generalizable Deepfake Detection via Consistency-Driven Frequency Debiasing (CVPR 2025 · similarity 0.82)
- CAV-MAE Sync: Improving Contrastive Audio-Visual Mask Autoencoders via Fine-Grained Alignment (CVPR 2025 · similarity 0.81)
- Robust Audio-Visual Segmentation via Audio-Guided Visual Convergent Alignment (CVPR 2025 · similarity 0.81)
- Object-aware Sound Source Localization via Audio-Visual Scene Understanding (CVPR 2025 · similarity 0.81)
- Seeing Sound, Hearing Sight: Uncovering Modality Bias and Conflict of AI models in Sound Localization (NeurIPS 2025 · similarity 0.81)
- D^3: Scaling Up Deepfake Detection by Learning from Discrepancy (CVPR 2025 · similarity 0.81)
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.
