3D Visual Grounding on Nr3D (val)
72.5Easy ScoreSceneVerse
Evaluation results
| Method | Links | |||||||
|---|---|---|---|---|---|---|---|---|
| SceneVerseSupervision=Fully Supervised2026.05 | 72.5 | 57.8 | 56.9 | 67.9 | 64.9 | — | — | |
| ViL3DRelSupervision Level=Fully-Supervised, Venue=Not specified2025.12 | 70.2 | 57.4 | 62 | 64.5 | 64.4 | — | — | |
| ViL3DRelSupervision=Fully Supervised2026.05 | 70.2 | 57.4 | 62 | 64.5 | 64.4 | — | — | |
| MiKASASupervision Level=Fully-Supervised, Venue=CVPR’242025.12 | 69.7 | 59.4 | 65.4 | 64 | 64.4 | — | — | |
| MiKASASupervision=Fully Supervised2026.05 | 69.7 | 59.4 | 65.4 | 64 | 64.4 | — | — | |
| BUTD-DETRSupervision Level=Fully-Supervised, Venue=ECCV’222025.12 | 60.7 | 48.4 | 46 | 58 | 54.6 | — | — | |
| AgentGrounderSupervision=Zero-Shot2026.05 | 59.6 | 45.4 | 47.9 | 54.5 | 52.4 | — | — | |
| VoG-72BSupervision Level=Zero-Shot, Model Scale=72B2025.12 | 58.9 | 37.2 | 39.4 | 52.1 | 47.6 | — | — | |
| VLM-GrounderSupervision=Zero-Shot2026.05 | 55.2 | 39.5 | 45.8 | 49.4 | 48 | — | — | |
| SeeGroundSupervision=Zero-Shot2026.05 | 54.5 | 38.3 | 42.3 | 48.2 | 46.1 | — | — | |
| SeeGround-72BSupervision Level=Zero-Shot, Venue=CVPR’25, Model Scale=72B2025.12 | 51.3 | 35.6 | 38.8 | 45.5 | 43.1 | — | — | |
| 3DVG-TSupervision Level=Fully-Supervised, Venue=ICCV’212025.12 | 48.5 | 34.8 | 34.8 | 43.7 | 40.8 | — | — | |
| ZSVG3D-GPT4Supervision Level=Zero-Shot, Venue=CVPR’242025.12 | 46.5 | 31.7 | 36.8 | 40 | 39 | — | — | |
| ZSVG3DSupervision=Zero-Shot2026.05 | 46.5 | 31.7 | 36.8 | 40 | 39 | — | — | |
| InstanceReferSupervision Level=Fully-Supervised, Venue=ICCV’212025.12 | 46 | 31.8 | 34.5 | 41.9 | 38.8 | — | — | |
| TGNNSupervision Level=Fully-Supervised, Venue=AAAI’212025.12 | 44.2 | 30.6 | 35.8 | 38 | 37.3 | — | — | |
| ReferIt3DNetSupervision Level=Fully-Supervised, Venue=ECCV’202025.12 | 43.6 | 27.9 | 32.5 | 37.1 | 35.6 | — | — | |
| VoG-2BSupervision Level=Zero-Shot, Model Scale=2B2025.12 | 40.3 | 22.6 | 30.4 | 31.5 | 31.1 | — | — | |
| SeeGround-2BSupervision Level=Zero-Shot, Venue=CVPR’25, Model Scale=2B2025.12 | 31.8 | 16.6 | 22.8 | 24.5 | 23.9 | — | — | |
| WS-3DVGSupervision Level=Weakly-Supervised, Venue=ICCV’232025.12 | 27.3 | 18 | 21.6 | 22.9 | 22.5 | — | — | |
| 3D-VisTAw/o. LLM/VLM=true, w/o. PC=false, Zero-Shot=false2026.07 | — | — | — | — | — | 47.7 | 42.2 | |
| BUTD-DETRw/o. LLM/VLM=true, w/o. PC=false, Zero-Shot=false2026.07 | — | — | — | — | — | 43.3 | — | |
| G2TAMw/o. LLM/VLM=true, w/o. PC=true, Zero-Shot=true2026.07 | — | — | — | — | — | 51.4 | 45.8 | |
| InstanceReferw/o. LLM/VLM=true, w/o. PC=false, Zero-Shot=false2026.07 | — | — | — | — | — | 29.9 | — | |
| LanguageReferw/o. LLM/VLM=true, w/o. PC=false, Zero-Shot=false2026.07 | — | — | — | — | — | 28.6 | — | |
| SAT-2Dw/o. LLM/VLM=true, w/o. PC=false, Zero-Shot=false2026.07 | — | — | — | — | — | 31.7 | — | |
| VLM-Grounderw/o. LLM/VLM=false, w/o. PC=true, Zero-Shot=true2026.07 | — | — | — | — | — | 48 | — |