Reproduce “Venus-DefakerOne2: A Vision-Forensics Foundation Model with Broader Coverage” on OpenResearch
Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.
Universal app for Apple silicon and Intel. Requires macOS 11 or later.
About this paper
Chenfan Qu, Yachao Liang, Mingqi Fang, Jiangwei Xie, Zhenming Wang, Baoyue Zhang, Changjiang Jiang, Zheng Chang and 23 more
Venus-DefakerOne2 is a 2-billion-parameter multimodal model for detecting digital manipulation and physical spoofing in images and videos. Trained on about 50 million samples, it evaluates video clips and sampled frames, then selects the prediction with lower Shannon entropy. On RRVBench, it reaches 95.3% balanced accuracy, versus 69.8% for VideoVeritas; this is a benchmark result, not a guarantee for real-world videos.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.