Data Science & Analytics
Physical AI Defect Image Generation
The skill generates realistic defect images for PCBA, metal surface, and glass inspection using NVIDIA Cosmos AnomalyGen on OSMO. It automates cold‑start data creation, image‑edit augmentation, and inference labeling, enabling rapid first‑time asset setup, finetuning checkpoints, and deployment configuration. Engineers and quality assurance teams benefit from streamlined defect dataset creation and accurate real‑image labeling for anomaly detection.
- Publisher / source
- NVIDIA Skills (Verified)
- Upstream record
- NVIDIA/skills
- Version
- Version 3
- Published
- Publication review
- Approved for publication
- Content fingerprint
02b77b4e1eedf523422a4f17fb6dd143cbfc044eca624ee8bbc716099c508828