3D Scene Reconstruction on ScanNet V2 (test) benchmark leaderboard · SOTA2 Research

3D Scene Reconstruction on ScanNet V2 (test)

0.735AccuracyNeRF

Evaluation Results

Method	Links
NeRF2025.03		0.735	0.177	—	0.131	0.291	0.176
CVD2023.04		0.344	9.1	21.8	0.266	0.461	0.331
NeuralAngelo2025.03		0.245	0.272	—	0.274	0.311	0.292
Geo-Neus2025.03		0.236	0.206	—	0.282	0.313	0.291
3DVNet2023.04		0.221	7.7	14.9	0.506	0.545	0.52
MonoSDF2025.03		0.214	0.18	—	0.297	0.325	0.31
MVDNet2023.04		0.205	8.4	14.5	0.231	0.473	0.307
NeuS2025.03		0.179	0.208	—	0.313	0.275	0.291
DPSNet2023.04		0.177	8.1	12.9	0.272	0.497	0.349
GPMVS2023.04		0.162	7.9	12.1	0.335	0.533	0.408
PermutoSDF2025.03		0.143	0.219	—	0.448	0.209	0.285
COLMAP2023.04		0.135	6.9	10.2	0.505	0.634	0.558
NeRFPrior2025.03		0.133	0.12	—	0.439	0.429	0.433
DeepVMVS2023.04		0.117	7.6	9.7	0.451	0.558	0.496
Atlas2023.04		0.084	10.2	9.3	0.565	0.598	0.578
TF2023.04		0.078	9.9	8.9	0.648	0.547	0.591
Manhattan2025.03		0.072	0.068	—	0.621	0.586	0.602
SimRec2023.04		0.065	7.8	7.2	0.641	0.581	0.608
VORTX2023.04		0.062	8.2	7.2	0.688	0.607	0.644
VisFusion2023.04		0.055	10.5	8	0.695	0.527	0.598
NeuRec2023.04		0.054	12.8	9.1	0.684	0.479	0.562
NeuRIS2025.03		0.054	0.052	—	0.729	0.684	0.705
Go-Surf2025.03		0.048	0.021	—	0.88	0.894	0.887
MonoSDF*2025.03		0.042	0.049	—	0.76	0.707	0.732
NeRFPrior2025.03		0.037	0.042	—	0.799	0.766	0.782
NeRFPrior2025.03		0.027	0.02	—	0.931	0.928	0.93