Browse › Computer Vision › 3D Semantic Scene Completion › NYUv2
NYUv2 Benchmark (3D Semantic Scene Completion)
This task was introduced in "Semantic Scene Completion from a Single Depth Image" (https://arxiv.org/abs/1611.08974) at CVPR 2017 . The target is to infer the dense 3D voxelized semantic scene from an incompleted 3D input (e.g. point cloud, depth map) and an optional RGB image. A recent summary can be found in the paper "3D Semantic Scene Completion: a Survey" (https://arxiv.org/abs/2103.07466), published at IJCV 2021.
The archive carries no text for this table; the description above is the archive's text for the task 3D Semantic Scene Completion. archive 2025-07-28
Over time archive 2025-07-28
The chart needs JavaScript; the table below carries every value.
Not inferred: mIoU. Points are placed at the row's paper date; 28 of 28 rows carry one.
Results archive 2025-07-28
Archive rows end at the archive snapshot, 2025-07-28: no result published after that date is in this table. Rank is the archive's row order at that snapshot; not re-ranked here. Metric values are the archive's strings. Column headers sort the table in your browser; each row keeps its archive rank.
| Paper | Code | Ran Syntology | Report | |||||
|---|---|---|---|---|---|---|---|---|
| 1 | SG-SSC | 55.4 | – | Paper | Code | 2024 | linked, not harvested | report |
| 2 | SISNet | 52.4 | – | Paper | Code | 2021 | 3 of 3 ran · 0 unverified | report |
| 3 | SPAwN (SUNCG pretraining) | 49.9 | – | Paper | Code | 2021 | linked, not harvested | report |
| 4 | SPAwN | 48 | – | Paper | Code | 2021 | linked, not harvested | report |
| 5 | Point-Voxel Aggregation Network | 46 | – | Paper | – | 2021 | no code linked | report |
| 6 | CCPNet (SUNCG pretraining) | 41.3 | ✓ | Paper | – | 2019 | no code linked | report |
| 7 | 3DSketch | 41.1 | – | Paper | Code | 2020 | 1 of 3 ran · 2 unverified | report |
| 8 | ForkNet (SUNCG pretraining) | 37.1 | ✓ | Paper | – | 2019 | no code linked | report |
| 9 | IPF-SPCNet: Semantic point completion network for 3D semantic scene completion | 35.1 | – | Paper | – | 2020 | no code linked | report |
| 10 | Real-time semantic scene completion via feature aggregation and conditioned prediction | 34.4 | – | Paper | – | 2021 | no code linked | report |
| 11 | TS3D | 34.1 | – | Paper | – | 2018 | no code linked | report |
| 12 | EdgeNet (SUNCG pretraining) | 33.7 | ✓ | Paper | – | 2021 | no code linked | report |
| 13 | AIC-Net | 33.3 | – | Paper | Code | 2020 | linked, not harvested | report |
| 14 | AMFNet | 33 | – | Paper | – | 2020 | no code linked | report |
| 15 | GRFNet | 32.9 | – | Paper | – | 2020 | no code linked | report |
| 16 | VD-CRF: Semantic scene completion with dense CRF from a single depth image. (SUNCG pretraining) | 31.8 | ✓ | Paper | – | 2021 | no code linked | report |
| 17 | Am2fnet: Attention-based multiscale & multi-modality fused network | 31.7 | – | Paper | – | 2021 | no code linked | report |
| 18 | OccDepth | 30.9 | – | Paper | Code | 2023 | 3 of 6 ran · 3 unverified | report |
| 19 | SSCNet (SUNCG pretraining) | 30.5 | ✓ | Paper | Code | 2016 | 1 of 1 ran · 0 unverified | report |
| 20 | DDRNet | 30.4 | – | Paper | – | 2019 | no code linked | report |
| 21 | VVNet | 29.3 | – | Paper | – | 2018 | no code linked | report |
| 22 | LMSCNet-SS | 28.4 | – | Paper | Code | 2020 | linked, not harvested | report |
| 23 | EdgeNet | 27.8 | – | Paper | Code | 2019 | linked, not harvested | report |
| 24 | Guedes et al. | 27.5 | – | Paper | – | 2018 | no code linked | report |
| 25 | MonoScene (RGB input only) | 26.94 | – | Paper | Code | 2021 | 3 of 6 ran · 3 unverified | report |
| 26 | ESSCNet | 26.7 | – | Paper | Code | 2019 | linked, not harvested | report |
| 27 | SSCNet | 24.7 | – | Paper | Code | 2016 | 1 of 1 ran · 0 unverified | report |
| 28 | 3D semantic scene completion from a single depth image using adversarial training | 22.7 | – | Paper | – | 2021 | no code linked | report |
All 28 rows shown. 28 link to a paper page on this site; 5 are marked as using additional training data in the archive. No GitHub stars are tracked; "Code" is the first repository the archive lists for the row. The archive carries no row tags, review links or community-submitted rows for this table; none are shown. archive 2025-07-28
Syntology Ran reads "N of M ran · U unverified": of the M code samples Syntology harvested from repositories linked to that row's paper (joined by arXiv id), N executed on a synthesized input and the other U = M−N are unverified (harvested, no recorded run). It counts code from repositories linked to that row's paper, not this result: the row's number was not reproduced and nothing here is a correctness claim. The other cell texts mean no graph line for the row: "linked, not harvested" (the archive links code, Syntology has not harvested it), "no code linked" (no code link in the archive), "not matched" (the row's paper URL matched no paper on this site). 6 rows have a graph line, from 5 distinct papers; 6 rows (5 papers) have at least one sample that ran. Counting each paper once: Syntology ran 11 of 19 samples; 8 unverified. Separately, 4 of those 19 are pointer-only (licence): the site points at that code rather than redistributing it, a licence property recorded for ran and unverified samples alike; each cell's tooltip carries the row's own pointer-only count. Read from the graph 2026-09-24. Per-sample status is on the paper page.
Report a problem or propose a change · a person checks every report against the paper or source before anything changes; decisions are listed on /corrections