Dissect and Prune: Enhancing Robustness in AI-Generated Image Detection
Quick Answer
This paper shows that The DEAR (Dissect and Prune) method improves AI-generated image detection by addressing prediction asymmetry, enhancing robustness against unseen generators and post-processing.
Quick Take
The DEAR (Dissect and Prune) method improves AI-generated image detection by addressing prediction asymmetry, enhancing robustness against unseen generators and post-processing. By pruning spurious features, DEAR retains only those that capture genuine artifacts, significantly boosting detection performance.
Key Points
- DEAR leverages inpainted images to prune distracting features in AI models.
- Features aligned with inpainted regions show less robustness to post-processing.
- The approach significantly mitigates prediction asymmetry in image detection.
- Experimental results indicate improved performance against unseen generators.
- Code for DEAR is publicly available on GitHub.
Paper Resources
Article Excerpt
From source RSS / original summaryarXiv:2606. 10309v1 Announce Type: new Abstract: While existing AI-generated image detectors report high performance, we identify that this is largely driven by a critical prediction asymmetry: a bias toward the real class that severely limits sensitivity to generated content, especially under standard post-processing operations such as compression and resizing. We hypothesize that this stems from the model's reliance on spurious features, distracting signals that obscure true generative artifacts.
To address this, we propose DEAR (Dissect and Prune), which leverages inpainted images to identify and prune these interfering components. Specifically, we find that features strongly aligned to either inpainted or non-inpainted regions are less robust to post-processing. By measuring the alignment between channel activations and inpaint masks, DEAR removes features at both extremes, retaining only those that capture genuine generative artifacts.
Experimental results demonstrate that our approach significantly enhances robustness against unseen generators and post-processing, effectively mitigating the prediction asymmetry. Our code is available at https://github. com/dahyedahye/dear.
Reader Mode unavailable (could not extract clean content).
Want this in your inbox every morning?
Daily brief at your local 8am — bilingual EN/中文, free.
More from arXiv cs.CV
See more →LLM-Guided ANN Index Optimization for Human-Object Interaction Retrieval
A phase-aware LLM agent optimizes human-object interaction retrieval, outperforming Optuna TPE by 33.3% and VDTuner by 34.2% on the HICO-DET benchmark. This method enhances throughput by 15.3x over UniIR and demonstrates strong transferability across vector database management systems.