NVIDIA applies VLMs and VFMs to semiconductor defect classification
Applied VLMs and VFMs to semiconductor defect classification — PCB from 93.84% to 98.51%
Why it matters
A combination of about 600 labeled images and domain adaptation on about 1 million unlabeled PCB images. The label efficiency has direct implications for the OSAT floor
Source — Primary sources first; secondary reports are labelled “Press”
Officialdeveloper.nvidia.comhttps://developer.nvidia.com/blog/optimizing-semiconductor-defect-classification-with-generative-ai-and-vision-foundation-models/Added in Week 35, 2026 · See that week