Scene Completion on NYU v2 (test)
78.2mIoUSISNet (DeepLabv3)
Evaluation results
| Method | Links | ||||
|---|---|---|---|---|---|
| SISNet (DeepLabv3)Resolution=(60, 60)2021.04 | 78.2 | 92.1 | 83.8 | — | |
| SISNet (BiSeNet)Resolution=(60, 60)2021.04 | 77.8 | 90.7 | 84.6 | — | |
| SketchResolution=(60, 60)2021.04 | 71.3 | 85 | 81.6 | — | |
| Baseline (BiSeNet)Resolution=(60, 60)2021.04 | 71 | 87.6 | 78.9 | — | |
| Baseline (DeepLabv3)Resolution=(60, 60)2021.04 | 70.8 | 88.7 | 77.7 | — | |
| CCPNetResolution=(240, 240)2021.04 | 63.5 | 74.2 | 90.8 | — | |
| ForkNetResolution=(80, 80)2021.04 | 63.4 | — | — | — | |
| VVNetR-120Resolution=(120, 60)2021.04 | 61.1 | 69.8 | 83.1 | — | |
| VVNetR-120train=SUNCG + NYU, input=d2019.08 | 61.1 | 69.8 | 83.1 | — | |
| DDR-Net2019.03 | 61 | 71.5 | 80.8 | — | |
| DDRNet2020.03 | 61 | 71.5 | 80.8 | — | |
| DDRNetResolution=(60, 60)2021.04 | 61 | 71.5 | 80.8 | — | |
| Garbade et al.train=SUNCG + NYU, input=d+s2019.08 | 60.7 | 69.5 | 82.7 | — | |
| SNetFusetrain=SUNCG + NYU, input=d+s2019.08 | 60.7 | 67.6 | 85.9 | — | |
| TNetFusetrain=SUNCG + NYU, input=d+s2019.08 | 60.7 | 67.3 | 85.8 | — | |
| SATNetResolution=(60, 60)2021.04 | 60.6 | 67.3 | 85.8 | — | |
| TS3DResolution=(240, 60)2021.04 | 60 | — | — | — | |
| AICNetResolution=(60, 60)2021.04 | 59.2 | 62.4 | 91.8 | — | |
| AMFNetconfiguration=Full model2020.03 | 59 | 67.9 | 82.3 | — | |
| AMFNetconfiguration=without attention2020.03 | 58.6 | 64.5 | 86.5 | — | |
| AMFNetconfiguration=basic architecture2020.03 | 58 | 64 | 86.6 | — | |
| EdgeNet-EFtrain=SUNCG + NYU, input=d+e2019.08 | 57.9 | 77 | 70 | — | |
| EdgeNet-LFtrain=SUNCG + NYU, input=d+e2019.08 | 57.9 | 77.6 | 69.5 | — | |
| AMFNetconfiguration=without segmentation guidance2020.03 | 57.3 | 68.5 | 78.6 | — | |
| AMFNetconfiguration=segmentation ground truth2020.03 | 57.2 | 66.3 | 80.5 | — | |
| EdgeNet-MFtrain=SUNCG + NYU, input=d+e2019.08 | 56.7 | 79.1 | 66.6 | — | |
| SSCNettrain=SUNCG + NYU, input=d2019.08 | 56.6 | 59.3 | 92.9 | — | |
| Guedes et al.train=SUNCG + NYU, input=d+c2019.08 | 56.6 | — | — | — | |
| EsscNet2019.03 | 56.2 | 71.9 | 71.9 | — | |
| Efficient Semantic Scene Completion Network2019.07 | 56.2 | 71.9 | 71.9 | — | |
| EsscNet2020.03 | 56.2 | 71.9 | 71.9 | — | |
| ESSCNetResolution=(240, 60)2021.04 | 56.2 | 71.9 | 71.9 | — | |
| EdgeNet-MFtrain=NYU, input=d+e2019.08 | 56.1 | 76 | 68.3 | — | |
| EdgeNet-LFtrain=NYU, input=d+e2019.08 | 55.4 | 75.5 | 67.5 | — | |
| SSCNet2019.03 | 55.1 | 57 | 94.5 | — | |
| SSCNet2019.07 | 55.1 | 57 | 94.5 | — | |
| SSCNet2020.03 | 55.1 | 57 | 94.5 | — | |
| SSCNetResolution=(240, 60)2021.04 | 55.1 | 57 | 94.5 | — | |
| SSCNettrain=NYU, input=d2019.08 | 55.1 | 57 | 94.5 | — | |
| EdgeNet-EFtrain=NYU, input=d+e2019.08 | 55.1 | 78.1 | 65.1 | — | |
| EdgeNet-EFtrain=SUNCG, input=d+e2019.08 | 53.6 | 61.9 | 80 | — | |
| SSCNettrain=SUNCG, input=d2019.08 | 53.2 | 55.6 | 91.9 | — | |
| EdgeNet-MFtrain=SUNCG, input=d+e2019.08 | 52.8 | 60.7 | 80.3 | — | |
| EdgeNet-LFtrain=SUNCG, input=d+e2019.08 | 52.3 | 59.9 | 80.5 | — | |
| Geiger et al.2019.03 | 44.4 | 65.7 | 58 | — | |
| Geiger et al.2020.03 | 44.4 | 65.7 | 58 | — | |
| Lin et al.2019.03 | 36.4 | 58.5 | 49.9 | — | |
| Lin et al.2020.03 | 36.4 | 58.5 | 49.9 | — | |
| 3DSketchVenue=CVPR’202026.03 | — | — | — | 38.64 | |
| AdaSFormerVenue=Ours2026.03 | — | — | — | 51.33 | |
| AICNetVenue=CVPR’202026.03 | — | — | — | 30.03 | |
| GenFuSEVenue=arXiv’252026.03 | — | — | — | 46.3 | |
| ISOVenue=ECCV’242026.03 | — | — | — | 47.11 | |
| LMSCNetVenue=3DV’202026.03 | — | — | — | 33.93 | |
| MonoMRNVenue=ICCV’252026.03 | — | — | — | 53.16 | |
| MonoSceneVenue=CVPR’222026.03 | — | — | — | 42.51 | |
| NDC-SceneVenue=ICCV’232026.03 | — | — | — | 44.17 |