Researchers report a test-time fine-tuning approach that adapts large-scale vision foundation models to segment electronic ...