# 3D Object Recognition on ScanObjectNN PB\_T50\_RS fine-tuned

Source: [3D Object Recognition on ScanObjectNN PB\_T50\_RS fine-tuned](<https://www.sota2.com/research/sota/3d-object-recognition-on-scanobjectnn-pb_t50_rs-fine-tuned>)

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

Data updated: 2026-02-26T05:54:25.003Z

Evaluation of 3D Object Recognition performance on the fine-tuned version of the ScanObjectNN PB\_T50\_RS dataset, typically measured by Overall Accuracy.

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 Object Recognition |
| Dataset (including any stated split) | ScanObjectNN PB\_T50\_RS fine-tuned |
| Metric | Overall Accuracy |
| Unit | PERCENT |
| Direction | Higher is better |
| Results loaded | 19 of 19 |
| 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

- [Overall Accuracy](<https://www.sota2.com/research/sota/3d-object-recognition-on-scanobjectnn-pb_t50_rs-fine-tuned.md?metric=Overall%20Accuracy>): Higher is better; unit: PERCENT.

## Recorded results

| Method | Overall Accuracy | Result date | Report paper |
| --- | --- | --- | --- |
| RECON++-L | 95.25 | Not recorded | 68fa740c1e209131cea5f73a |
| PointGPT-L | 93.4 | Not recorded | 68fa740c1e209131cea5f73a |
| RECON++-B | 93.34 | Not recorded | 68fa740c1e209131cea5f73a |
| PointGPT-B | 91.9 | Not recorded | 68fa740c1e209131cea5f73a |
| ULIP-2 | 91.5 | Not recorded | 68fa740c1e209131cea5f73a |
| RECON | 91.26 | Not recorded | 68fa740c1e209131cea5f73a |
| I2P-MAE | 90.11 | Not recorded | 68fa740c1e209131cea5f73a |
| VPP | 89.28 | Not recorded | 68fa740c1e209131cea5f73a |
| TAP | 88.5 | Not recorded | 68fa740c1e209131cea5f73a |
| ACT | 88.21 | Not recorded | 68fa740c1e209131cea5f73a |
| PointNeXt | 87.7 | Not recorded | 68fa740c1e209131cea5f73a |
| Point2Vec | 87.5 | Not recorded | 68fa740c1e209131cea5f73a |
| Point-M2AE | 86.43 | Not recorded | 68fa740c1e209131cea5f73a |
| PointMLP | 85.4 | Not recorded | 68fa740c1e209131cea5f73a |
| Point-MAE | 85.18 | Not recorded | 68fa740c1e209131cea5f73a |
| Point-BERT | 83.07 | Not recorded | 68fa740c1e209131cea5f73a |
| DGCNN | 78.1 | Not recorded | 68fa740c1e209131cea5f73a |
| PointNet++ | 77.9 | Not recorded | 68fa740c1e209131cea5f73a |
| PointNet | 68 | Not recorded | 68fa740c1e209131cea5f73a |

### RECON++-L

Overall Accuracy: 95.25. Result ID: 699316803a6c31484913a1dd.

[Reporting paper](<https://www.sota2.com/research/paper/68fa740c1e209131cea5f73a.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa740c1e209131cea5f73a.md>)

[Implementation](<https://qizekun.github.io/ShapeLLM>)

Recorded evaluation configuration: {"Training Paradigm":"Self-Supervised","Post-pretraining":true,"Voting":true}. This record does not assert independent reproduction.

### PointGPT-L

Overall Accuracy: 93.4. Result ID: 699316803a6c31484913a1db.

[Reporting paper](<https://www.sota2.com/research/paper/68fa740c1e209131cea5f73a.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6f671e209131cea455b3.md>)

[Implementation](<https://github.com/CGuangyan-BIT/PointGPT>)

Recorded evaluation configuration: {"Training Paradigm":"Self-Supervised","Post-pretraining":true,"Voting":true}. This record does not assert independent reproduction.

### RECON++-B

Overall Accuracy: 93.34. Result ID: 699316803a6c31484913a1dc.

[Reporting paper](<https://www.sota2.com/research/paper/68fa740c1e209131cea5f73a.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa740c1e209131cea5f73a.md>)

[Implementation](<https://qizekun.github.io/ShapeLLM>)

Recorded evaluation configuration: {"Training Paradigm":"Self-Supervised","Post-pretraining":true,"Voting":true}. This record does not assert independent reproduction.

### PointGPT-B

Overall Accuracy: 91.9. Result ID: 699316803a6c31484913a1da.

[Reporting paper](<https://www.sota2.com/research/paper/68fa740c1e209131cea5f73a.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6f671e209131cea455b3.md>)

[Implementation](<https://github.com/CGuangyan-BIT/PointGPT>)

Recorded evaluation configuration: {"Training Paradigm":"Self-Supervised","Post-pretraining":true,"Voting":true}. This record does not assert independent reproduction.

### ULIP-2

Overall Accuracy: 91.5. Result ID: 699316803a6c31484913a1d8.

[Reporting paper](<https://www.sota2.com/research/paper/68fa740c1e209131cea5f73a.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa730c1e209131cea5a192.md>)

[Implementation](<https://github.com/salesforce/ULIP>)

Recorded evaluation configuration: {"Training Paradigm":"Self-Supervised","Post-pretraining":false,"Voting":true}. This record does not assert independent reproduction.

### RECON

Overall Accuracy: 91.26. Result ID: 699316803a6c31484913a1d9.

[Reporting paper](<https://www.sota2.com/research/paper/68fa740c1e209131cea5f73a.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6f721e209131cea45a6b.md>)

[Implementation](<https://github.com/qizekun/ReCon>)

Recorded evaluation configuration: {"Training Paradigm":"Self-Supervised","Post-pretraining":false,"Voting":true}. This record does not assert independent reproduction.

### I2P-MAE

Overall Accuracy: 90.11. Result ID: 699316803a6c31484913a1d7.

[Reporting paper](<https://www.sota2.com/research/paper/68fa740c1e209131cea5f73a.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6e291e209131cea3e3b3.md>)

[Implementation](<https://github.com/ZrrSkywalker/I2P-MAE>)

Recorded evaluation configuration: {"Training Paradigm":"Self-Supervised","Post-pretraining":false,"Voting":true}. This record does not assert independent reproduction.

### VPP

Overall Accuracy: 89.28. Result ID: 699316803a6c31484913a1d6.

[Reporting paper](<https://www.sota2.com/research/paper/68fa740c1e209131cea5f73a.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa70e91e209131cea4de6b.md>)

[Implementation](<https://github.com/qizekun/VPP>)

Recorded evaluation configuration: {"Training Paradigm":"Self-Supervised","Post-pretraining":false,"Voting":true}. This record does not assert independent reproduction.

### TAP

Overall Accuracy: 88.5. Result ID: 699316803a6c31484913a1d5.

[Reporting paper](<https://www.sota2.com/research/paper/68fa740c1e209131cea5f73a.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa707c1e209131cea4b669.md>)

[Implementation](<https://github.com/wangzy22/TAP>)

Recorded evaluation configuration: {"Training Paradigm":"Self-Supervised","Post-pretraining":false,"Voting":true}. This record does not assert independent reproduction.

### ACT

Overall Accuracy: 88.21. Result ID: 699316803a6c31484913a1d4.

[Reporting paper](<https://www.sota2.com/research/paper/68fa740c1e209131cea5f73a.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6e741e209131cea3ff1b.md>)

[Implementation](<https://github.com/RunpeiDong/ACT>)

Recorded evaluation configuration: {"Training Paradigm":"Self-Supervised","Post-pretraining":false,"Voting":true}. This record does not assert independent reproduction.

### PointNeXt

Overall Accuracy: 87.7. Result ID: 699316803a6c31484913a1cf.

[Reporting paper](<https://www.sota2.com/research/paper/68fa740c1e209131cea5f73a.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6d9f1e209131cea3aee1.md>)

[Implementation](<https://github.com/guochengqian/pointnext>)

Recorded evaluation configuration: {"Training Paradigm":"Supervised","Post-pretraining":false,"Voting":true}. This record does not assert independent reproduction.

### Point2Vec

Overall Accuracy: 87.5. Result ID: 699316803a6c31484913a1d3.

[Reporting paper](<https://www.sota2.com/research/paper/68fa740c1e209131cea5f73a.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa70d71e209131cea4d949.md>)

Recorded evaluation configuration: {"Training Paradigm":"Self-Supervised","Post-pretraining":false,"Voting":true}. This record does not assert independent reproduction.

### Point-M2AE

Overall Accuracy: 86.43. Result ID: 699316803a6c31484913a1d2.

[Reporting paper](<https://www.sota2.com/research/paper/68fa740c1e209131cea5f73a.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6daa1e209131cea3b524.md>)

[Implementation](<https://github.com/ZrrSkywalker/Point-M2AE>)

Recorded evaluation configuration: {"Training Paradigm":"Self-Supervised","Post-pretraining":false,"Voting":true}. This record does not assert independent reproduction.

### PointMLP

Overall Accuracy: 85.4. Result ID: 699316803a6c31484913a1ce.

[Reporting paper](<https://www.sota2.com/research/paper/68fa740c1e209131cea5f73a.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6e061e209131cea3d7c8.md>)

[Implementation](<https://github.com/ma-xu/pointMLP-pytorch>)

Recorded evaluation configuration: {"Training Paradigm":"Supervised","Post-pretraining":false,"Voting":true}. This record does not assert independent reproduction.

### Point-MAE

Overall Accuracy: 85.18. Result ID: 699316803a6c31484913a1d1.

[Reporting paper](<https://www.sota2.com/research/paper/68fa740c1e209131cea5f73a.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6c1b1e209131cea32397.md>)

[Implementation](<https://github.com/Pang-Yatian/Point-MAE>)

Recorded evaluation configuration: {"Training Paradigm":"Self-Supervised","Post-pretraining":false,"Voting":true}. This record does not assert independent reproduction.

### Point-BERT

Overall Accuracy: 83.07. Result ID: 699316803a6c31484913a1d0.

[Reporting paper](<https://www.sota2.com/research/paper/68fa740c1e209131cea5f73a.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6cba1e209131cea35a33.md>)

[Implementation](<https://github.com/lulutang0608/Point-BERT>)

Recorded evaluation configuration: {"Training Paradigm":"Self-Supervised","Post-pretraining":false,"Voting":true}. This record does not assert independent reproduction.

### DGCNN

Overall Accuracy: 78.1. Result ID: 699316803a6c31484913a1cd.

[Reporting paper](<https://www.sota2.com/research/paper/68fa740c1e209131cea5f73a.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa65e11e209131cea0d1ac.md>)

[Implementation](<https://github.com/WangYueFt/dgcnn>)

Recorded evaluation configuration: {"Training Paradigm":"Supervised","Post-pretraining":false,"Voting":true}. This record does not assert independent reproduction.

### PointNet++

Overall Accuracy: 77.9. Result ID: 699316803a6c31484913a1cc.

[Reporting paper](<https://www.sota2.com/research/paper/68fa740c1e209131cea5f73a.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa63641e209131ce9fe5a5.md>)

Recorded evaluation configuration: {"Training Paradigm":"Supervised","Post-pretraining":false,"Voting":true}. This record does not assert independent reproduction.

### PointNet

Overall Accuracy: 68. Result ID: 699316803a6c31484913a1cb.

[Reporting paper](<https://www.sota2.com/research/paper/68fa740c1e209131cea5f73a.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa634f1e209131ce9fe019.md>)

Recorded evaluation configuration: {"Training Paradigm":"Supervised","Post-pretraining":false,"Voting":true}. This record does not assert independent reproduction.

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