Research map / Vision-language and multimodal / Cross-modal learning
Unified Frequency-Assisted Transformer Framework for Detecting and Grounding Multi-modal Manipulation
Huan Liu, Zichang Tan, Qiang Chen, Yunchao Wei, Yao Zhao, Jingdong Wang
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 frequency encoder (ablation) | UFAFormer · DGM4 val | IoU_m: 76.33 -> 77.82 | Improved |
| add frequency encoder (ablation) | UFAFormer · DGM4 val | Text grounding F1: 68.75 -> 70.33 | Improved |
| add forgery-aware mutual module (ablation) | UFAFormer · DGM4 val | IoU_m: 74.72 -> 77.82 | Improved |
| add unified decoder instead of standard transformer decoder (ablation) | UFAFormer · DGM4 val | AUC: 92.33 -> 93.41 | Improved |
| add unified decoder instead of standard transformer decoder (ablation) | UFAFormer · DGM4 val | mAP: 85.13 -> 86.98 | Improved |
| add FAFP and MCA together instead of either alone (ablation) | UFAFormer · DGM4 val | IoU_m: 76.57 -> 77.82 | Improved |
| use intra- and inter-band attention instead of standard self-attention (ablation) | UFAFormer · DGM4 val | IoU_m: 76.78 -> 77.82 | Improved |
| add inter-band interaction to intra-band only (ablation) | UFAFormer · DGM4 val | IoU_m: 75.48 -> 77.82 | Improved |
| add forgery aggregation by prepended sub-band embedding (ablation) | UFAFormer · DGM4 val | IoU_m: 75.83 -> 77.82 | Improved |
| add forgery-aware selection sampling rates (2,6,8 vs 10,10,10) (ablation) | UFAFormer · DGM4 val | IoU_m: 74.02 -> 77.82 | Improved |
| increase frequency encoder layers from 3 to 4 (ablation) | UFAFormer · DGM4 val | IoU_m: 77.82 -> 77.69 | No clear effect |
| increase grounding embeddings from 5 to 7 (ablation) | UFAFormer · DGM4 val | IoU_m: 77.82 -> 77.89 | No clear effect |
Closest papers by meaning
- IDseq: Decoupled and Sequentially Detecting and Grounding Multi-Modal Media Manipulation (AAAI 2025 · similarity 0.86)
- Unleashing the Potential of Consistency Learning for Detecting and Grounding Multi-Modal Media Manipulation (CVPR 2025 · similarity 0.86)
- Towards General Visual-Linguistic Face Forgery Detection (CVPR 2025 · similarity 0.86)
- Identity-Aware Vision-Language Model for Explainable Face Forgery Detection (AAAI 2026 · similarity 0.86)
- Rethinking Vision-Language Model in Face Forensics: Multi-Modal Interpretable Forged Face Detector (CVPR 2025 · similarity 0.85)
- On Learning Multi-Modal Forgery Representation for Diffusion Generated Video Detection (NeurIPS 2024 · similarity 0.85)
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.
