AI Paper Atlas

Research map / Vision-language and multimodal / Multimodal LLMs

AIGI-Holmes: Towards Explainable and Generalizable AI-Generated Image Detection via Multimodal Large Language Models

Ziyin Zhou, Yunpeng Luo, Yuanchen Wu, Ke Sun, Jiayi Ji, Ke Yan, Shouhong Ding, Xiaoshuai Sun, Yunsheng Wu, Rongrong Ji

ICCV 2025 · Vision-language and multimodal › Multimodal LLMs · approach: Interleaved image-text generation with MLLMs · 46 citations (OpenAlex)

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 changedWhereMeasuredEffect
add Visual Expert Pre-training Stage (VEP-S) (ablation)LLaVA-based AIGI-Holmes · P1 / P3 mean accuracyaccuracy: 83.3 -> 86.8 (P1); 90.1 -> 97.2 (P3)Improved
add DPO to baseline LLaVA (ablation)AIGI-Holmes · P1 / P3accuracy: 83.3 -> 84.8; 90.1 -> 92.3Improved
add DPO on top of VEP-S (ablation)AIGI-Holmes · P1 / P3accuracy: 86.8 -> 87.4; 97.2 -> 97.6Improved
add Collaborative Decoding on top of VEP-S (ablation)AIGI-Holmes · P1 / P3accuracy: 86.8 -> 90.8; 97.2 -> 98.9Improved
add DPO and CD on top of VEP-S (ablation)AIGI-Holmes · P1 / P3accuracy: 87.4 -> 93.2; 97.6 -> 99.2Improved
add DPO stage after SFT (ablation)AIGI-Holmes explanations · Explanation qualityhuman ELO rating: 10.670 -> 11.420Improved
add DPO stage after SFT (ablation)AIGI-Holmes explanations · Explanation qualityBLEU-1: 0.445 -> 0.622Improved
add NPR expert alongside CLIP visual expert (ablation)AIGI-Holmes · P1 / P3accuracy: 92.0 -> 93.2; 98.9 -> 99.2Improved
add CLIP visual expert over LLaVA-only encoder (ablation)AIGI-Holmes · P1 / P3accuracy: 86.8 -> 92.0; 98.2 -> 98.9Improved
use Holmes pipeline instead of LoRA on LLM only (ablation)AIGI-Holmes · P1 / P3accuracy: 83.3 -> 86.8; 90.1 -> 98.2Improved
use Holmes pipeline instead of LoRA on all modules (ablation)AIGI-Holmes · P1 / P3accuracy: 84.8 -> 86.8; 91.2 -> 98.2Improved
use Mistral-7B instead of Llama3-8B as LLM (ablation)AIGI-Holmes · P1 / P3accuracy: 91.5 -> 93.2; 97.8 -> 99.2Improved

Later papers that compared with this method

2 later accepted papers report AIGI-Holmes as a baseline in their results tables; 2 of them report a better number on most of the results they share with it.

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

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.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy