DINOv3 + SAM 3.1 Turns a Surgical Tray into a Local Visual Search Engine

MaziyarPanahi · x · 2026-10-03

Combining DINOv3 with SAM 3.1, the author turned a surgical instrument tray into a visual search engine: one instrument serves as the query, and the rest are ranked by visual similarity—entirely on-device, fully local. A striking demonstration of how specialized vision model combos keep improving.

Related event: Dev Combines DINOv3 and SAM 3.1 into Fully Local Visual Search Engine(2 posts)→

Original post →

More from Models

Models channel →