CoordRefer: Coordinate-Aware 3D Visual Grounding from Multiview ImagesPublished in arXiv 2026, 2026Share on Twitter Facebook LinkedIn Previous Next