Paper Detail

Venus-DefakerOne2: A Vision-Forensics Foundation Model with Broader Coverage

Chenfan Qu, Yachao Liang, Mingqi Fang, Jiangwei Xie, Zhenming Wang, Baoyue Zhang, Changjiang Jiang, Zheng Chang, Fanli Jin, Xuekang Zhu, Kaiwen Feng, Junchi Li, Weihong Zhang, Haoyue Wang, Chenfeng Zhang, Yinan Xia, Fangming Cui, Dong Fan, Jingjing Liu, Tianyi Zhang, Luobin Chen, Ji-Zhe Zhou, Sheng Li, Wei Wang, Bin Li, Ran He, Xinpeng Zhang, Linna Zhou, Jian Liu, Weiqiang Wang

huggingface Score 10.0

Published 2026-10-27 · First seen 2026-11-02

Research Track A · General AI

Abstract

Real-world forgery is inherently heterogeneous: it manifests not only as digital manipulation but also as physical spoofing, and is carried not only by images but also by videos. Despite substantial progress, a unified anti-forgery foundation model covering such a broad threat spectrum remains absent. In this paper, we present Venus-DefakerOne2, which substantially extends the forensic capability boundary along both the forgery-form and media-type dimensions. Trained on tens of millions of open-source and in-house samples, it incorporates two key designs. Tide-Style-Tuning alternates LoRA and full-parameter fine-tuning to balance computational efficiency, generalization, and hard-sample fitting while mitigating catastrophic forgetting. Entropy-based Test Adaptation determines the optimal prediction modality by analyzing prediction entropy. Venus-DefakerOne2 achieves new state-of-the-art performance across a comprehensive suite of benchmarks, including DeepFake faces (96.7%), AI-generated images (93.3%), tampered documents (87.6%), manipulated natural images (89.6%), physical spoofs (93.1%), and AI-generated videos (96.0%). It significantly outperforms both specialized expert models and its predecessor, DefakerOne, establishing itself as a powerful and versatile foundation model for binary-decision forensic tasks.

Workflow Status

Review status
pending
Role
unreviewed
Read priority
soon
Vote
Not set.
Saved
no
Collections
Not filed yet.
Next action
Not filled yet.

Reading Brief

No structured notes yet. Add `summary_sections`, `why_relevant`, `claim_impact`, or `next_action` in `papers.jsonl` to enrich this view.

Why It Surfaced

No ranking explanation is available yet.

Tags

No tags.

BibTeX

@misc{qu2026venus,
  title = {Venus-DefakerOne2: A Vision-Forensics Foundation Model with Broader Coverage},
  author = {Chenfan Qu and Yachao Liang and Mingqi Fang and Jiangwei Xie and Zhenming Wang and Baoyue Zhang and Changjiang Jiang and Zheng Chang and Fanli Jin and Xuekang Zhu and Kaiwen Feng and Junchi Li and Weihong Zhang and Haoyue Wang and Chenfeng Zhang and Yinan Xia and Fangming Cui and Dong Fan and Jingjing Liu and Tianyi Zhang and Luobin Chen and Ji-Zhe Zhou and Sheng Li and Wei Wang and Bin Li and Ran He and Xinpeng Zhang and Linna Zhou and Jian Liu and Weiqiang Wang},
  year = {2026},
  abstract = {Real-world forgery is inherently heterogeneous: it manifests not only as digital manipulation but also as physical spoofing, and is carried not only by images but also by videos. Despite substantial progress, a unified anti-forgery foundation model covering such a broad threat spectrum remains absent. In this paper, we present Venus-DefakerOne2, which substantially extends the forensic capability boundary along both the forgery-form and media-type dimensions. Trained on tens of millions of open-},
  url = {https://huggingface.co/papers/2610.37918},
  keywords = {LoRA, full-parameter fine-tuning, catastrophic forgetting, entropy-based test adaptation, foundation model, code available, huggingface daily},
  eprint = {2610.37918},
  archiveprefix = {arXiv},
}

Metadata

{}