3D Object Detection on ScanNet (val)
77.7mAP@0.25UniGeo
Evaluation results
| Method | Links | ||||||||||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| UniGeoVenue=-, Aggregation=Best Results2026.01 | 77.7 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 65.6 | |
| UniDet3DVenue=AAAI’2025, Aggregation=Best Results2026.01 | 77.5 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 66.1 | |
| V-DETRVenue=ICLR’2024, Aggregation=Best Results2026.01 | 77.4 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 65 | |
| UniDet3DVenue=AAAI’2025, Aggregation=Average across 25 trials2026.01 | 77.1 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 65.2 | |
| V-DETRVenue=ICLR’2024, Aggregation=Average across 25 trials2026.01 | 76.8 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 64.5 | |
| UniGeoVenue=-, Aggregation=Average across 25 trials2026.01 | 76.8 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 64.5 | |
| SWIN3D-S + CAGroup3DSupervision Type=Based on pretraining2023.04 | 76.4 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 62.7 | |
| SWIN3D-L + CAGroup3DSupervision Type=Based on pretraining2023.04 | 76.4 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 63.2 | |
| CAGroup3DRepresentation Type=Point Cloud2024.10 | 75.12 | — | 60.4 | 93 | 95.3 | 92.3 | 69.9 | 67.9 | 63.6 | 67.3 | 40.7 | 77 | 83.9 | 69.4 | 65.7 | 73 | 100 | 79.7 | 87 | 66.1 | — | — | |
| CAGroup3DSupervision Type=Supervised2023.04 | 75.1 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 61.3 | |
| SPGroup3DVenue=AAAI’2024, Aggregation=Best Results2026.01 | 74.3 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 59.6 | |
| SWIN3D-S + FCAF3DSupervision Type=Based on pretraining2023.04 | 74.2 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 59.5 | |
| SWIN3D-L + FCAF3DSupervision Type=Based on pretraining2023.04 | 74.2 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 58.6 | |
| SPGroup3DVenue=AAAI’2024, Aggregation=Average across 25 trials2026.01 | 73.5 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 58.3 | |
| SWIN3D-S* + CAGroup3DSupervision Type=Supervised2023.04 | 73.3 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 58.6 | |
| TR3DVenue=ICIP’2023, Aggregation=Best Results2026.01 | 72.9 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 59.3 | |
| SWIN3D-S* + FCAF3DSupervision Type=Supervised2023.04 | 72.1 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 56.8 | |
| TR3DVenue=ICIP’2023, Aggregation=Average across 25 trials2026.01 | 72 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 57.4 | |
| SoftGroupSupervision Type=Supervised2023.04 | 71.6 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 59.4 | |
| FCAF3D2021.12 | 71.5 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 8 | 57.3 | |
| FCAF3DSupervision Type=Supervised2023.04 | 71.5 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 57.3 | |
| FCAF3DVenue=ECCV’2022, Aggregation=Best Results2026.01 | 71.5 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 57.3 | |
| FCAF3DRepresentation Type=Point Cloud2024.10 | 71.5 | — | 57.2 | 87 | 95 | 92.3 | 70.3 | 61.1 | 60.2 | 64.5 | 29.9 | 64.3 | 71.5 | 60.1 | 52.4 | 83.9 | 99.9 | 84.7 | 86.6 | 65.4 | — | — | |
| RepSurfSupervision Type=Supervised2023.04 | 71.2 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 54.8 | |
| SOFWVenue=TMM’2025, Aggregation=Best Results2026.01 | 70.9 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 52.3 | |
| FCAF3DVenue=ECCV’2022, Aggregation=Average across 25 trials2026.01 | 70.7 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 56 | |
| RBGNetNo. of class obtained=18, Supervision=Supervised2022.10 | 70.6 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| FCAF3DFeature levels=32021.12 | 69.8 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 12.2 | 53.6 | |
| GroupFree2021.12 | 69.1 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 6.6 | 52.8 | |
| RandomRoomsSupervision Type=Based on pretraining2023.04 | 68.6 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 51.5 | |
| H3DNet2021.12 | 67.2 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 4.9 | 48.1 | |
| BRNet2021.12 | 66.1 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 10.3 | 50.9 | |
| 3DETR-m2021.12 | 65 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 3.1 | 47 | |
| Ponder-RGBDDetection Model=VoteNet, Pre-training Type=Rendering, Pre-training Epochs=1002023.10 | 63.6 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 41 | |
| MaskPointDetection Model=3DETR, Pre-training Type=Masked Auto-Encoding, Pre-training Epochs=3002023.10 | 63.4 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 40.6 | |
| FCAF3DFeature levels=22021.12 | 63.1 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 31.5 | 46.8 | |
| GSDN2021.12 | 62.8 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 20.1 | 34.8 | |
| 3DETR2021.12 | 62.7 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 3.1 | 37.5 | |
| 3DETRDetection Model=3DETR2023.10 | 62.7 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 37.5 | |
| DepthContrastDetection Model=VoteNet, Pre-training Type=Contrast, Pre-training Epochs=10002023.10 | 62.1 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 39.1 | |
| PC-FractalDBDetection Model=VoteNet, Pre-training Type=Contrast2023.10 | 61.9 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 38.3 | |
| IAEDetection Model=VoteNet, Pre-training Type=Masked Auto-Encoding, Pre-training Epochs=10002023.10 | 61.5 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 39.8 | |
| RandomRoomsDetection Model=VoteNet, Pre-training Type=Contrast, Pre-training Epochs=3002023.10 | 61.3 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 36.2 | |
| Point-BERTDetection Model=3DETR, Pre-training Type=Masked Auto-Encoding, Pre-training Epochs=3002023.10 | 61 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 38.3 | |
| 3DGS-DETRepresentation Type=View-synthesis2024.10 | 59.9 | — | 44.1 | 82.7 | 81.7 | 79.6 | 56 | 35.4 | 27.6 | 45.2 | 17.3 | 61.9 | 72.8 | 40.7 | 56.6 | 71.9 | 98.5 | 72.2 | 88.3 | 46.7 | — | — | |
| STRLDetection Model=VoteNet, Pre-training Type=Contrast, Pre-training Epochs=1002023.10 | 59.5 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 38.4 | |
| PointContrastSupervision Type=Based on pretraining2023.04 | 59.2 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 37.3 | |
| PointContrastDetection Model=VoteNet, Pre-training Type=Contrast2023.10 | 59.2 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 38 | |
| VoteNet (Fast Point Transformer backbone)Backbone=Fast Point Transformer2021.12 | 59.1 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 35.6 | |
| VoteNetRepresentation Type=Point Cloud2024.10 | 58.7 | — | 36.3 | 87.9 | 88.7 | 89.6 | 58.8 | 47.3 | 38.1 | 44.6 | 7.8 | 56.1 | 71.7 | 47.2 | 45.4 | 57.1 | 94.9 | 54.7 | 92.1 | 37.2 | — | — | |
| VoteNet2021.12 | 58.6 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 11.8 | 33.5 | |
| VoteNetNo. of class obtained=18, Supervision=Supervised2022.10 | 58.6 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| VoteNetDetection Model=VoteNet2023.10 | 58.6 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 33.5 | |
| CN-RMARepresentation Type=Multi-view Image2024.10 | 58.6 | — | 42.3 | 80 | 79.4 | 83.1 | 55.2 | 44 | 30.6 | 53.6 | 8.8 | 65 | 70 | 44.9 | 44 | 55.2 | 95.4 | 68.1 | 86.1 | 49.7 | — | — | |
| VoteNet (MinkowskiNet† backbone)Backbone=MinkowskiNet, input point sub-sampling=none2021.12 | 55.3 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 33 | |
| ImGeoNetRepresentation Type=Multi-view Image2024.10 | 54.6 | — | 40.6 | 84.1 | 74.8 | 75.6 | 59.9 | 40.4 | 24.7 | 60.1 | 4.2 | 41.2 | 70.9 | 33.7 | 54.4 | 47.5 | 95.2 | 57.5 | 81.5 | 36.1 | — | — | |
| VoteNet (PointNet++ backbone)Backbone=PointNet++2021.12 | 54.2 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 30.1 | |
| NeRF-Det++Representation Type=View-synthesis2024.10 | 53.9 | — | 36.1 | 82.9 | 74.9 | 79.1 | 57 | 37.3 | 24.9 | 54.6 | 2.4 | 51.7 | 72.2 | 25.5 | 58.7 | 51.5 | 92.7 | 50.8 | 82.2 | 35.1 | — | — | |
| VoteNet (MinkowskiNet backbone)Backbone=MinkowskiNet2021.12 | 53.8 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 30.2 | |
| NeRF-DetRepresentation Type=View-synthesis2024.10 | 53.3 | — | 37.6 | 84.9 | 76.2 | 76.7 | 57.5 | 36.4 | 17.8 | 47 | 2.5 | 49.2 | 52 | 29.2 | 68.2 | 49.3 | 97.1 | 57.6 | 83.6 | 35.9 | — | — | |
| VoteNet (RS-CNN backbone)Backbone=RS-CNN2021.12 | 51.6 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 29.5 | |
| Point-M2AESupervision Type=Based on pretraining2023.04 | 50.1 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 33.2 | |
| ImVoxelNetRepresentation Type=Multi-view Image2024.10 | 49 | — | 30.9 | 84 | 77.5 | 73.3 | 56.7 | 35.1 | 18.6 | 47.5 | 0 | 44.4 | 65.5 | 19.6 | 58.2 | 32.8 | 92.3 | 40.1 | 77.6 | 28 | — | — | |
| VoteNet (KPConv backbone)Backbone=KPConv2021.12 | 48.9 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | 29.2 | |
| VoteNet + BRPNo. of class obtained=18, Supervision=Weakly-supervised2022.10 | 31.2 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| GSPNNo. of class obtained=18, Supervision=Supervised2022.10 | 30.6 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| F-PointNetNo. of class obtained=18, Supervision=Supervised2022.10 | 19.8 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| WyPR + priorNo. of class obtained=18, Supervision=Weakly-supervised2022.10 | 19.7 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| VoteNet + WS3DNo. of class obtained=18, Supervision=Weakly-supervised2022.10 | 18.4 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| WyPRNo. of class obtained=18, Supervision=Weakly-supervised2022.10 | 18.3 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| MIL-detNo. of class obtained=18, Supervision=Weakly-supervised2022.10 | 9.6 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| SL3DSubscript=1600, No. of class obtained=182022.10 | 9.3 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| SL3DSubscript=200, No. of class obtained=152022.10 | 7.9 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| SL3DSubscript=100, No. of class obtained=122022.10 | 7.6 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| SL3DSubscript=50, No. of class obtained=72022.10 | 4.6 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| 3D-SISRGB=false, PC=true2021.06 | — | 25.4 | 12.8 | 63.1 | 66 | 46.3 | 26.9 | 8 | 2.8 | 2.3 | 0 | 6.9 | 33.3 | 2.5 | 10.4 | 12.2 | 74.5 | 22.9 | 58.7 | 7.1 | — | — | |
| 3D-SISRGB=true, PC=true2021.06 | — | 40.2 | 19.8 | 69.7 | 66.2 | 71.8 | 36.1 | 30.6 | 10.9 | 27.3 | 0 | 10 | 46.9 | 14.1 | 53.8 | 36 | 87.6 | 43 | 84.3 | 16.2 | — | — | |
| 3DSIS2021.05 | — | 40.2 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| F-PointNet2021.05 | — | 10.8 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| GSPN2021.05 | — | 17.7 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| H3DNetRGB=false, PC=true2021.06 | — | 67.2 | 49.4 | 88.6 | 91.8 | 90.2 | 64.9 | 61 | 51.9 | 54.9 | 18.6 | 62 | 75.9 | 57.3 | 57.2 | 75.3 | 97.9 | 67.4 | 92.5 | 53.6 | — | — | |
| ImVoxelNetRGB=true, PC=false2021.06 | — | 48.1 | 28.5 | 84.4 | 73.1 | 70.1 | 51.9 | 32.2 | 15 | 34.2 | 1.6 | 29.7 | 66.1 | 23.5 | 57.8 | 43.2 | 92.4 | 54.1 | 74 | 34.9 | — | — | |
| MIL-det (unsup. GSS)#boxes=<1k2021.05 | — | 9.6 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| VoteNet#boxes=2562021.05 | — | 58.6 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| VoteNet#boxes=1k2021.05 | — | 55.3 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| VoteNetRGB=false, PC=true2021.06 | — | 58.7 | 36.3 | 87.9 | 88.7 | 89.6 | 58.8 | 47.3 | 38.1 | 44.6 | 7.8 | 56.1 | 71.7 | 47.2 | 45.4 | 57.1 | 94.9 | 54.7 | 92.1 | 37.2 | — | — | |
| WyPR#boxes=<1k2021.05 | — | 18.3 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | |
| WyPR+prior#boxes=<1k, External Object Prior=Per-class mean shapes from synthetic datasets2021.05 | — | 19.7 | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — | — |