ResearchBenchmarksVisual Grounding on Multi3DRefer (ZT/ST/MT Variants)Follow66.3ZT (w/o D)V3DM1.71618.48335.2552.017Jun 18, 2026Evaluation ResultsMethodMethodLinksZT (w/o D)ZT (w/ D)ST (w/o D)ST (w/ D)MTOverall AccuracyV3DMTrained=ViGiL3D++SR, D...Trained=ViGiL3D++SR, Detector=Mask3D2026.0666.342.122.91214.218.2V3DMTrained=ViGiL3D++, Det...Trained=ViGiL3D++, Detector=Mask3D2026.0651.535.717.57.38.512.63D-VisTATrained=ScanScribe, De...Trained=ScanScribe, Detector=Mask3D2026.0620.110.823.311.211.814GPSTrained=SceneVerse, De...Trained=SceneVerse, Detector=Mask3D2026.064.21.926.514.416.516.3