CALIPER: Metric-Grounded Model-Free Recognition of Visually Similar Industrial Parts

CALIPER, a model-free RGB-D framework for recognizing visually similar industrial parts, achieved 88.2% closed-set accuracy with 99.8% localization recall across 18 parts, according to an arXiv paper (arXiv:2609.17820v1). The system combines a YOLOv8n-seg localizer, a frozen DINOv2 backbone with an episodically trained embedding head, and margin-conditioned metric fusion that improved unseen-class accuracy by up to 37.4 percentage points. After 10-shot enrollment of two held-out screws, CALIPER reached 85.7% overall accuracy, and a robot-arm deployment identified 17 of 18 parts without deployment-specific retraining.

arXiv:2609.17820v1 Announce Type: new Abstract: Fine-grained recognition of visually similar industrial parts is challenging when classes differ primarily in physical dimensions. Normalizing detected object crops to a fixed input size suppresses absolute scale, while CAD models and large class-specific datasets may be unavailable in evolving industrial inventories. We present CALIPER, a model-free RGB-D framework that couples support-based appearance matching with metric size evidence. Each training class is onboarded from a single turntable RGB-D video and one to two labeled real images; 3D reconstruction provides novel-view appearance support, while aligned depth yields a class-specific metric size profile. At inference, a coarse YOLOv8n-seg model localizes parts, and a frozen DINOv2 backbone with an episodically trained embedding head performs fine-grained support matching. Margin-conditioned metric fusion activates probabilistic size evidence only for appearance-ambiguous decisions. New classes are enrolled from a small RGB-D support set without updating network parameters. We evaluate CALIPER on 18 visually similar industrial parts: 16 classes are used for training, while two screws are reserved for training-free enrollment. CALIPER achieves 88.2% closed-set accuracy with 99.8% localization recall and 85.7% overall accuracy after 10-shot enrollment of the two unseen screws. Metric fusion improves unseen-class accuracy by up to 37.4 percentage points without statistically significant degradation of the original inventory. Robot-arm deployment identifies 17/18 parts without deployment-specific retraining.