Davidlequnchen/VLM-CADFeatureRecognition

Employs prompt engineering techniques—multi-view isometric rendering, few-shot learning, and chain-of-thought reasoning—to enable zero-shot generalization across diverse manufacturing features without predefined geometric rules. Integrates PyTorch with pythonocc for CAD geometry processing and STEP file parsing, paired with the MFCAD-VLM dataset of expert-annotated CAD models at varying complexity levels to benchmark VLM performance on feature detection tasks.