Research map / Vision-language and multimodal / Robot action and navigation
Forensic Prompting with Dual-Action Policy Optimization for Vision-Language Forgery Detection and Localization
Ye Zhu, Ai Zhao, Jinwei 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 Forensic Prompting Module (prompt bank + router) (ablation) | FPDA base · CocoGlide detection | AUC: 0.905 -> 0.925 | Improved |
| add DAPO image-side action (ablation) | FPDA base · CocoGlide localization | F1: 0.690 -> 0.730 | Improved |
| combine FPM and DAPO (ablation) | FPDA · CocoGlide detection | AUC: 0.925 -> 0.941 | Improved |
| add Prompt Bank (ablation) | FPM · SID Set description detection | AUC: 0.985 -> 0.989 | Improved |
| add Prompt Router (ablation) | FPM · SID Set description detection | AUC: 0.989 -> 0.993 | Improved |
| fuse optional text without reliability gating (ablation) | FPM bank+router · SID Set description detection | AUC: 0.993 -> 0.992 | Hurt |
| add reliability gate to text fusion (ablation) | FPM · SID Set description detection | AUC: 0.992 -> 0.996 | Improved |
| enable text-side action (ablation) | DAPO · NIST16 detection | AUC: 0.831 -> 0.863 | Improved |
| enable image-side action (ablation) | DAPO · NIST16 detection | AUC: 0.831 -> 0.871 | Improved |
| use both text and image actions (ablation) | DAPO · NIST16 detection | AUC: 0.871 -> 0.902 | Improved |
| add detection reward (ablation) | DAPO · CASIA v1 detection | AUC: 0.953 -> 0.975 | Improved |
| add localization reward (ablation) | DAPO · CASIA v1 localization | F1: 0.702 -> 0.738 | Improved |
Earlier methods this paper compares with
Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 on most shared results.
- FakeShield — FakeShield: Explainable Image Forgery Detection and Localization via Multi-modal Large Language Models (ICLR 2025) · better on 1 of 1 shared result
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- Reinforced Multi-teacher Knowledge Distillation for Efficient General Image Forgery Detection and Localization (AAAI 2025 · similarity 0.86)
- Toward Real-world Text Image Forgery Localization: Structured and Interpretable Data Synthesis (NeurIPS 2025 · similarity 0.86)
- Unlocking the Capabilities of Large Vision-Language Models for Generalizable and Explainable Deepfake Detection (ICML 2025 · similarity 0.86)
- Identity-Aware Vision-Language Model for Explainable Face Forgery Detection (AAAI 2026 · similarity 0.85)
- OmniVL-Guard: Towards Unified Vision-Language Forgery Detection and Grounding via Balanced RL (ICML 2026 · similarity 0.85)
- Rethinking Vision-Language Model in Face Forensics: Multi-Modal Interpretable Forged Face Detector (CVPR 2025 · 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.
