Zero-Shot 3D Visual Grounding from Vision-Language Models Paper • 2505.22429 • Published May 28, 2025 • 2