New Vision-Forensics Foundation Model Targets Broader Range of Real-World Forgeries
A paper on arXiv introduces Venus-DefakerOne2, a vision-forensics foundation model designed to cover a broader range of heterogeneous forgeries, including both digital manipulation and physical spoofing.
Read the full story at arXiv cs.CV ↗
Timeline · 1 report
- 2026-10-30 04:00 · arXiv cs.CV
Venus-DefakerOne2: A Vision-Forensics Foundation Model with Broader Coverage