{"paper":{"title":"Reduce the Artifacts Bias for More Generalizable AI-Generated Image Detection","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"A GAN-based upsampling method plus Separate Expert Fusion reduces artifact bias and improves generalization in AI-generated image detection.","cross_cats":[],"primary_cat":"cs.CV","authors_text":"Gao Li, Wenhao Wang, Yang Yang, Yiheng Li, Zhen Lei, Zichang Tan","submitted_at":"2026-05-14T07:26:36Z","abstract_excerpt":"As the misuse of AI-generated images grows, generalizable image detection techniques are urgently needed. Recent state-of-the-art (SOTA) methods adopt aligned training datasets to reduce content, size, and format biases, empowering models to capture robust forgery cues. A common strategy is to employ reconstruction techniques, e.g., VAE and DDIM, which show remarkable results in diffusion-based methods. However, such reconstruction-based approaches typically introduce limited and homogeneous artifacts, which cannot fully capture diverse generative patterns, such as GAN-based methods. To comple"},"claims":{"count":4,"items":[{"kind":"strongest_claim","text":"Rather than merely benefiting GAN-generated image detection, this design introduces diverse and complementary artifact patterns that enable SEF to learn a more robust decision boundary and improve generalization across broader generative methods.","source":"verdict.strongest_claim","status":"machine_extracted","claim_id":"C1","attestation":"unclaimed"},{"kind":"weakest_assumption","text":"That the proposed GAN-based upsampling produces artifact patterns that are both aligned (content/size/format) with reconstruction fakes and sufficiently distinct to provide complementary information without introducing new unmodeled biases.","source":"verdict.weakest_assumption","status":"machine_extracted","claim_id":"C2","attestation":"unclaimed"},{"kind":"one_line_summary","text":"SEF introduces GAN upsampling for diverse artifacts and expert fusion to reduce domain interference, yielding stronger generalization on 13 benchmarks for AI-generated image detection.","source":"verdict.one_line_summary","status":"machine_extracted","claim_id":"C3","attestation":"unclaimed"},{"kind":"headline","text":"A GAN-based upsampling method plus Separate Expert Fusion reduces artifact bias and improves generalization in AI-generated image detection.","source":"verdict.pith_extraction.headline","status":"machine_extracted","claim_id":"C4","attestation":"unclaimed"}],"snapshot_sha256":"7fd062025d9d7bea7f3472a20f1aeafcd4874baabd153e283353af4addf22f26"},"source":{"id":"2605.14486","kind":"arxiv","version":1},"verdict":{"id":"b99a2f40-c57d-4514-bbcb-187aa876e820","model_set":{"reader":"grok-4.3"},"created_at":"2026-05-15T02:33:22.200340Z","strongest_claim":"Rather than merely benefiting GAN-generated image detection, this design introduces diverse and complementary artifact patterns that enable SEF to learn a more robust decision boundary and improve generalization across broader generative methods.","one_line_summary":"SEF introduces GAN upsampling for diverse artifacts and expert fusion to reduce domain interference, yielding stronger generalization on 13 benchmarks for AI-generated image detection.","pipeline_version":"pith-pipeline@v0.9.0","weakest_assumption":"That the proposed GAN-based upsampling produces artifact patterns that are both aligned (content/size/format) with reconstruction fakes and sufficiently distinct to provide complementary information without introducing new unmodeled biases.","pith_extraction_headline":"A GAN-based upsampling method plus Separate Expert Fusion reduces artifact bias and improves generalization in AI-generated image detection."},"references":{"count":68,"sample":[{"doi":"","year":2023,"title":"Synthbuster: Towards detection of diffusion model generated images.IEEE Open Journal of Signal Processing, 5:1–9, 2023","work_id":"2febba00-8ed5-42ff-a42c-45224344b8b9","ref_index":1,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":2018,"title":"Large Scale GAN Training for High Fidelity Natural Image Synthesis","work_id":"244e6f06-bad2-4f34-8186-ff370286427f","ref_index":2,"cited_arxiv_id":"1809.11096","is_internal_anchor":true},{"doi":"","year":2026,"title":"Zooming in on fakes: A novel dataset for localized ai-generated image detection with forgery amplification approach","work_id":"b34aecdd-ac2e-49c5-9124-c0c7fb332b03","ref_index":3,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":2021,"title":"Emerging properties in self-supervised vision transformers","work_id":"def40bfd-07b5-4f25-94fd-883b4ac6838d","ref_index":4,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":2024,"title":"Real-time deepfake detection in the real-world","work_id":"9493fae4-3b84-47d9-9b27-12baa60d298c","ref_index":5,"cited_arxiv_id":"","is_internal_anchor":false}],"resolved_work":68,"snapshot_sha256":"d2dfa4238ec4a34659b0762f58bedb56d162eed7f3bc6c56bca208b86708500e","internal_anchors":8},"formal_canon":{"evidence_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"author_claims":{"count":0,"strong_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"builder_version":"pith-number-builder-2026-05-17-v1"}