Kestrel: Point Grounding Multimodal LLM for Part-Aware 3D Vision-Language Understanding

Open in new window