# 3D Instance Segmentation on SUNCG synthetic scans

Source: [3D Instance Segmentation on SUNCG synthetic scans](<https://www.sota2.com/research/sota/3d-instance-segmentation-on-suncg-synthetic-scans>)

Collection index: [llms.txt](<https://www.sota2.com/research/llms.txt>)

Data updated: 2026-02-26T06:09:58.970Z

Evaluation of 3D Instance Segmentation performance on the SUNCG synthetic scans dataset, reporting Mean Average Precision at IoU=0.25 (mAP@0.25) and per-class metrics for common indoor objects.

A research result applies to its stated dataset, split, metric and evaluation configuration. Compare like-for-like settings, preserve the metric's unit and direction, and distinguish paper-reported results from independent reproduction. The best result recorded here is not a claim of exhaustive, current global coverage.

| Scope | Recorded value |
| --- | --- |
| Task | 3D Instance Segmentation |
| Dataset (including any stated split) | SUNCG synthetic scans |
| Metric | mAP@0.25 |
| Unit | PERCENT |
| Direction | Higher is better |
| Results loaded | 7 of 7 |
| Coverage | All results currently returned for this benchmark |

Evaluation settings may differ between results. The original papers and recorded configurations are needed to establish comparability. A shared metric name alone does not establish an identical protocol.

## Metrics

- [mAP@0.25](<https://www.sota2.com/research/sota/3d-instance-segmentation-on-suncg-synthetic-scans.md?metric=mAP%400.25>): Higher is better; unit: PERCENT.

- [AP@0.25 (Cabinet)](<https://www.sota2.com/research/sota/3d-instance-segmentation-on-suncg-synthetic-scans.md?metric=AP%400.25%20(Cabinet)>): Higher is better; unit: PERCENT.

- [AP@0.25 (Bed)](<https://www.sota2.com/research/sota/3d-instance-segmentation-on-suncg-synthetic-scans.md?metric=AP%400.25%20(Bed)>): Higher is better; unit: PERCENT.

- [AP@0.25 (Chair)](<https://www.sota2.com/research/sota/3d-instance-segmentation-on-suncg-synthetic-scans.md?metric=AP%400.25%20(Chair)>): Higher is better; unit: PERCENT.

- [AP@0.25 (Sofa)](<https://www.sota2.com/research/sota/3d-instance-segmentation-on-suncg-synthetic-scans.md?metric=AP%400.25%20(Sofa)>): Higher is better; unit: PERCENT.

- [AP@0.25 (Table)](<https://www.sota2.com/research/sota/3d-instance-segmentation-on-suncg-synthetic-scans.md?metric=AP%400.25%20(Table)>): Higher is better; unit: PERCENT.

- [AP@0.25 (Door)](<https://www.sota2.com/research/sota/3d-instance-segmentation-on-suncg-synthetic-scans.md?metric=AP%400.25%20(Door)>): Higher is better; unit: PERCENT.

- [AP@0.25 (Window)](<https://www.sota2.com/research/sota/3d-instance-segmentation-on-suncg-synthetic-scans.md?metric=AP%400.25%20(Window)>): Higher is better; unit: PERCENT.

- [AP@0.25 (Bookshelf)](<https://www.sota2.com/research/sota/3d-instance-segmentation-on-suncg-synthetic-scans.md?metric=AP%400.25%20(Bookshelf)>): Higher is better; unit: PERCENT.

- [AP@0.25 (Entrance)](<https://www.sota2.com/research/sota/3d-instance-segmentation-on-suncg-synthetic-scans.md?metric=AP%400.25%20(Entrance)>): Higher is better; unit: PERCENT.

- [AP@0.25 (Desk)](<https://www.sota2.com/research/sota/3d-instance-segmentation-on-suncg-synthetic-scans.md?metric=AP%400.25%20(Desk)>): Higher is better; unit: PERCENT.

- [AP@0.25 (Shelf)](<https://www.sota2.com/research/sota/3d-instance-segmentation-on-suncg-synthetic-scans.md?metric=AP%400.25%20(Shelf)>): Higher is better; unit: PERCENT.

- [AP@0.25 (Curtain)](<https://www.sota2.com/research/sota/3d-instance-segmentation-on-suncg-synthetic-scans.md?metric=AP%400.25%20(Curtain)>): Higher is better; unit: PERCENT.

- [AP@0.25 (Drawer)](<https://www.sota2.com/research/sota/3d-instance-segmentation-on-suncg-synthetic-scans.md?metric=AP%400.25%20(Drawer)>): Higher is better; unit: PERCENT.

- [AP@0.25 (Mirror)](<https://www.sota2.com/research/sota/3d-instance-segmentation-on-suncg-synthetic-scans.md?metric=AP%400.25%20(Mirror)>): Higher is better; unit: PERCENT.

- [AP@0.25 (TV)](<https://www.sota2.com/research/sota/3d-instance-segmentation-on-suncg-synthetic-scans.md?metric=AP%400.25%20(TV)>): Higher is better; unit: PERCENT.

- [AP@0.25 (Nightstand)](<https://www.sota2.com/research/sota/3d-instance-segmentation-on-suncg-synthetic-scans.md?metric=AP%400.25%20(Nightstand)>): Higher is better; unit: PERCENT.

- [AP@0.25 (Toilet)](<https://www.sota2.com/research/sota/3d-instance-segmentation-on-suncg-synthetic-scans.md?metric=AP%400.25%20(Toilet)>): Higher is better; unit: PERCENT.

- [AP@0.25 (Sink)](<https://www.sota2.com/research/sota/3d-instance-segmentation-on-suncg-synthetic-scans.md?metric=AP%400.25%20(Sink)>): Higher is better; unit: PERCENT.

- [AP@0.25 (Lamp)](<https://www.sota2.com/research/sota/3d-instance-segmentation-on-suncg-synthetic-scans.md?metric=AP%400.25%20(Lamp)>): Higher is better; unit: PERCENT.

- [AP@0.25 (Bathtub)](<https://www.sota2.com/research/sota/3d-instance-segmentation-on-suncg-synthetic-scans.md?metric=AP%400.25%20(Bathtub)>): Higher is better; unit: PERCENT.

- [AP@0.25 (Other Structure)](<https://www.sota2.com/research/sota/3d-instance-segmentation-on-suncg-synthetic-scans.md?metric=AP%400.25%20(Other%20Structure)>): Higher is better; unit: PERCENT.

- [AP@0.25 (Other Furniture)](<https://www.sota2.com/research/sota/3d-instance-segmentation-on-suncg-synthetic-scans.md?metric=AP%400.25%20(Other%20Furniture)>): Higher is better; unit: PERCENT.

- [AP@0.25 (Other Object)](<https://www.sota2.com/research/sota/3d-instance-segmentation-on-suncg-synthetic-scans.md?metric=AP%400.25%20(Other%20Object)>): Higher is better; unit: PERCENT.

## Recorded results

| Method | mAP@0.25 | Result date | Report paper |
| --- | --- | --- | --- |
| 3D-SIS | 30.6 | Not recorded | 68fa65ac1e209131cea0bda7 |
| 3D-SIS | 29.4 | Not recorded | 68fa65ac1e209131cea0bda7 |
| 3D-SIS | 29.1 | Not recorded | 68fa65ac1e209131cea0bda7 |
| 3D-SIS | 26.8 | Not recorded | 68fa65ac1e209131cea0bda7 |
| SGPN | 22.5 | Not recorded | 68fa65ac1e209131cea0bda7 |
| Mask R-CNN | 19.9 | Not recorded | 68fa65ac1e209131cea0bda7 |
| Seg-Cluster | 16.4 | Not recorded | 68fa65ac1e209131cea0bda7 |

### 3D-SIS

mAP@0.25: 30.6. Result ID: 698cc30e8b517eb4331c8fc8.

[Reporting paper](<https://www.sota2.com/research/paper/68fa65ac1e209131cea0bda7.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa65ac1e209131cea0bda7.md>)

[Implementation](<https://github.com/Sekunde/3D-SIS>)

Recorded evaluation configuration: {"Input features":"geometry + color","Number of views":"5"}. This record does not assert independent reproduction.

### 3D-SIS

mAP@0.25: 29.4. Result ID: 698cc30e8b517eb4331c8fc7.

[Reporting paper](<https://www.sota2.com/research/paper/68fa65ac1e209131cea0bda7.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa65ac1e209131cea0bda7.md>)

[Implementation](<https://github.com/Sekunde/3D-SIS>)

Recorded evaluation configuration: {"Input features":"geometry + color","Number of views":"3"}. This record does not assert independent reproduction.

### 3D-SIS

mAP@0.25: 29.1. Result ID: 698cc30e8b517eb4331c8fc6.

[Reporting paper](<https://www.sota2.com/research/paper/68fa65ac1e209131cea0bda7.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa65ac1e209131cea0bda7.md>)

[Implementation](<https://github.com/Sekunde/3D-SIS>)

Recorded evaluation configuration: {"Input features":"geometry + color","Number of views":"1"}. This record does not assert independent reproduction.

### 3D-SIS

mAP@0.25: 26.8. Result ID: 698cc30e8b517eb4331c8fc5.

[Reporting paper](<https://www.sota2.com/research/paper/68fa65ac1e209131cea0bda7.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa65ac1e209131cea0bda7.md>)

[Implementation](<https://github.com/Sekunde/3D-SIS>)

Recorded evaluation configuration: {"Input features":"geometry only"}. This record does not assert independent reproduction.

### SGPN

mAP@0.25: 22.5. Result ID: 698cc30e8b517eb4331c8fc4.

[Reporting paper](<https://www.sota2.com/research/paper/68fa65ac1e209131cea0bda7.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa65db1e209131cea0cde1.md>)

[Implementation](<https://www.sota2.com/github.com/laughtervv/SGPN>)

Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction.

### Mask R-CNN

mAP@0.25: 19.9. Result ID: 698cc30e8b517eb4331c8fc3.

[Reporting paper](<https://www.sota2.com/research/paper/68fa65ac1e209131cea0bda7.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa63f21e209131cea01b22.md>)

[Implementation](<https://github.com/facebookresearch/Detectron>)

Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction.

### Seg-Cluster

mAP@0.25: 16.4. Result ID: 698cc30e8b517eb4331c8fc2.

[Reporting paper](<https://www.sota2.com/research/paper/68fa65ac1e209131cea0bda7.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa65db1e209131cea0cde1.md>)

[Implementation](<https://www.sota2.com/github.com/laughtervv/SGPN>)

Recorded evaluation configuration: Not recorded. This record does not assert independent reproduction.

Page 1 of 1. This is an index page, not the full collection.
