Research map / Image generation / Autoregressive and latent generation
RiSE: Residual Subspace Expert for Generalizable Text-Centric Image Forgery Localization
Kahim Wong, Kemou Li, Yiming Chen, Haiwei Wu, Jiantao Zhou
Read the paperCheck your idea against 137,217 papers
AI-extracted summary. Written by AI from the paper's full text. Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Proposes RiSE, a residual subspace expert approach for text-centric image forgery localization that aims to generalize to unseen forgery types better than full-parameter fine-tuning.
Problem
Text-centric image forgery localization (TFL) detectors rely on full-parameter fine-tuning, which causes low-rank feature collapse and overfitting to training artifacts, hurting generalization to unseen forgery types.
Approach
Uses a residual subspace expert formulation as an alternative to full-parameter fine-tuning to preserve feature rank and reduce overfitting to training artifacts; details beyond the abstract are not available.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| replace full-parameter fine-tuning with residual subspace adaptation (ablation) | RiSE-9 backbone · Avg F1 (7 test sets) | Avg F1: 0.353 -> 0.625 | Improved |
| increase number of residual experts from 1 to 9 (ablation) | RiSE · Cross-domain avg F1 | AvgC F1: 0.542 -> 0.611 | Improved |
| increase queries from 1 to 10 or 20 (ablation) | RiSE-9 · Avg F1 (7 test sets) | Avg F1: 0.625 -> 0.418 (10 queries), 0.469 (20 queries) | Hurt |
| make the mask module shared across experts (ablation) | RiSE-9 · Avg F1 (7 test sets) | Avg F1: 0.579 -> 0.625 | Improved |
| use offline expert assignment instead of learned MoE routing (ablation) | RiSE-9 · Avg F1 (7 test sets) | Avg F1: 0.603 -> 0.625 (standard MoE), 0.541 -> 0.625 (two-stage MoE) | Improved |
| adapt both attention and MLP layers rather than one (ablation) | RiSE-9 · Avg F1 (7 test sets) | Avg F1: 0.590 (attention only), 0.612 (MLP only) -> 0.625 | Improved |
| use higher residual rank (4, 8, 16) instead of rank 1 (ablation) | RiSE-9 · Avg F1 (7 test sets) | Avg F1: 0.625 -> 0.624 / 0.616 / 0.621 | No clear effect |
| use LoRA-style adaptation instead of SVD residual adaptation (ablation) | RiSE-9 · Avg F1 (7 test sets) | Avg F1: 0.625 -> 0.617 to 0.623 | Hurt |
| always fuse DCT features during training (p_g=1) rather than random gating (ablation) | RiSE-9 with DCT · Cross-domain avg F1 | AvgC F1: 0.596 -> 0.519 | Hurt |
| add random-gated DCT fusion (p_g=0.5) (ablation) | RiSE-9 · In-domain avg F1 | AvgI F1: 0.663 -> 0.690 | Improved |
| add DCT features to RiSE-9 | RiSE-9 · Cross-domain avg F1 | AvgC F1: 0.611 -> 0.616 | No clear effect |
| use latent perturbation confidence (LPC) instead of SDC for expert selection (ablation) | RiSE-9 · Cross-domain avg F1 | AvgC F1: 0.534 -> 0.611 | 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.
- ADCD-Net — ADCD-Net: Robust Document Image Forgery Localization via Adaptive DCT Feature and Hierarchical Content Disentanglement (ICCV 2025) · better on 4 of 5 shared results
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.
Experimental setup
Baselines: full-parameter fine-tuning TFL detectors
Closest papers by meaning
- Toward Real-world Text Image Forgery Localization: Structured and Interpretable Data Synthesis (NeurIPS 2025 · similarity 0.87)
- FakeShield: Explainable Image Forgery Detection and Localization via Multi-modal Large Language Models (ICLR 2025 · similarity 0.86)
- Reinforced Multi-teacher Knowledge Distillation for Efficient General Image Forgery Detection and Localization (AAAI 2025 · similarity 0.86)
- A Lottery Ticket Hypothesis Approach with Sparse Fine-tuning and MAE for Image Forgery Detection and Localization (AAAI 2025 · similarity 0.86)
- SUMI-IFL: An Information-Theoretic Framework for Image Forgery Localization with Sufficiency and Minimality Constraints (AAAI 2025 · similarity 0.86)
- Revisiting Tampered Scene Text Detection in the Era of Generative AI (AAAI 2025 · similarity 0.86)
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.
