# Primitive Fitting on Primitive Fitting (Out-of-distribution)

Source: [Primitive Fitting on Primitive Fitting (Out-of-distribution)](<https://www.sota2.com/research/sota/primitive-fitting-on-primitive-fitting-out-of-distribution>)

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

Data updated: 2026-02-26T06:10:15.903Z

Evaluation of Primitive Fitting performance on the Primitive Fitting (Out-of-distribution) dataset, measured by mIoU.

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 | Primitive Fitting |
| Dataset (including any stated split) | Primitive Fitting (Out-of-distribution) |
| Metric | mIoU |
| Unit | PERCENT |
| Direction | Higher is better |
| Results loaded | 14 of 14 |
| 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

- [mIoU](<https://www.sota2.com/research/sota/primitive-fitting-on-primitive-fitting-out-of-distribution.md?metric=mIoU>): Higher is better; unit: PERCENT.

## Recorded results

| Method | mIoU | Result date | Report paper |
| --- | --- | --- | --- |
| AutoGPart (HPNet) | 80.4 | Not recorded | 68fa6c5a1e209131cea337d2 |
| HPNet | 79.5 | Not recorded | 68fa6c5a1e209131cea337d2 |
| AutoGPart (PN++) | 76.5 | Not recorded | 68fa6c5a1e209131cea337d2 |
| AutoGPart (DGCNN) | 73.4 | Not recorded | 68fa6c5a1e209131cea337d2 |
| SPFN | 72.3 | Not recorded | 68fa6c5a1e209131cea337d2 |
| PointNet++ | 71.6 | Not recorded | 68fa6c5a1e209131cea337d2 |
| Gradient Surgery (DGCNN) | 71.6 | Not recorded | 68fa6c5a1e209131cea337d2 |
| MixStyle (DGCNN) | 71.4 | Not recorded | 68fa6c5a1e209131cea337d2 |
| MixStyle (PN++) | 70.9 | Not recorded | 68fa6c5a1e209131cea337d2 |
| Gradient Surgery (PN++) | 70.3 | Not recorded | 68fa6c5a1e209131cea337d2 |
| HPNet\* | 69.6 | Not recorded | 68fa6c5a1e209131cea337d2 |
| Meta-learning (DGCNN) | 69.3 | Not recorded | 68fa6c5a1e209131cea337d2 |
| Meta Learning (PN++) | 68.5 | Not recorded | 68fa6c5a1e209131cea337d2 |
| DGCNN | 68 | Not recorded | 68fa6c5a1e209131cea337d2 |

### AutoGPart (HPNet)

mIoU: 80.4. Result ID: 698f08d73a6c314849f51c17.

[Reporting paper](<https://www.sota2.com/research/paper/68fa6c5a1e209131cea337d2.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6c5a1e209131cea337d2.md>)

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

Recorded evaluation configuration: {"Segmentation Module Type":"Clustering-based"}. This record does not assert independent reproduction.

### HPNet

mIoU: 79.5. Result ID: 698f08d73a6c314849f51c14.

[Reporting paper](<https://www.sota2.com/research/paper/68fa6c5a1e209131cea337d2.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6b0d1e209131cea2bbda.md>)

Recorded evaluation configuration: {"Segmentation Module Type":"Clustering-based"}. This record does not assert independent reproduction.

### AutoGPart (PN++)

mIoU: 76.5. Result ID: 698f08d73a6c314849f51c15.

[Reporting paper](<https://www.sota2.com/research/paper/68fa6c5a1e209131cea337d2.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6c5a1e209131cea337d2.md>)

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

Recorded evaluation configuration: {"Backbone":"PointNet++","Segmentation Module Type":"Classification-based"}. This record does not assert independent reproduction.

### AutoGPart (DGCNN)

mIoU: 73.4. Result ID: 698f08d73a6c314849f51c16.

[Reporting paper](<https://www.sota2.com/research/paper/68fa6c5a1e209131cea337d2.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6c5a1e209131cea337d2.md>)

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

Recorded evaluation configuration: {"Backbone":"DGCNN","Segmentation Module Type":"Classification-based"}. This record does not assert independent reproduction.

### SPFN

mIoU: 72.3. Result ID: 698f08d73a6c314849f51c12.

[Reporting paper](<https://www.sota2.com/research/paper/68fa6c5a1e209131cea337d2.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa67001e209131cea139f6.md>)

Recorded evaluation configuration: {"Segmentation Module Type":"Classification-based"}. This record does not assert independent reproduction.

### PointNet++

mIoU: 71.6. Result ID: 698f08d73a6c314849f51c0a.

[Reporting paper](<https://www.sota2.com/research/paper/68fa6c5a1e209131cea337d2.md>)

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

Recorded evaluation configuration: {"Backbone":"PointNet++","Segmentation Module Type":"Classification-based"}. This record does not assert independent reproduction.

### Gradient Surgery (DGCNN)

mIoU: 71.6. Result ID: 698f08d73a6c314849f51c0d.

[Reporting paper](<https://www.sota2.com/research/paper/68fa6c5a1e209131cea337d2.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6b291e209131cea2c675.md>)

Recorded evaluation configuration: {"Backbone":"DGCNN","Training Strategy":"Gradient Surgery","Segmentation Module Type":"Classification-based"}. This record does not assert independent reproduction.

### MixStyle (DGCNN)

mIoU: 71.4. Result ID: 698f08d73a6c314849f51c11.

[Reporting paper](<https://www.sota2.com/research/paper/68fa6c5a1e209131cea337d2.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa69ac1e209131cea233f3.md>)

[Implementation](<https://github.com/KaiyangZhou/mixstyle-release>)

Recorded evaluation configuration: {"Backbone":"DGCNN","Training Strategy":"MixStyle","Segmentation Module Type":"Classification-based"}. This record does not assert independent reproduction.

### MixStyle (PN++)

mIoU: 70.9. Result ID: 698f08d73a6c314849f51c10.

[Reporting paper](<https://www.sota2.com/research/paper/68fa6c5a1e209131cea337d2.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa69ac1e209131cea233f3.md>)

[Implementation](<https://github.com/KaiyangZhou/mixstyle-release>)

Recorded evaluation configuration: {"Backbone":"PointNet++","Training Strategy":"MixStyle","Segmentation Module Type":"Classification-based"}. This record does not assert independent reproduction.

### Gradient Surgery (PN++)

mIoU: 70.3. Result ID: 698f08d73a6c314849f51c0c.

[Reporting paper](<https://www.sota2.com/research/paper/68fa6c5a1e209131cea337d2.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6b291e209131cea2c675.md>)

Recorded evaluation configuration: {"Backbone":"PointNet++","Training Strategy":"Gradient Surgery","Segmentation Module Type":"Classification-based"}. This record does not assert independent reproduction.

### HPNet\*

mIoU: 69.6. Result ID: 698f08d73a6c314849f51c13.

[Reporting paper](<https://www.sota2.com/research/paper/68fa6c5a1e209131cea337d2.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6b0d1e209131cea2bbda.md>)

Recorded evaluation configuration: {"Segmentation Module Type":"Clustering-based"}. This record does not assert independent reproduction.

### Meta-learning (DGCNN)

mIoU: 69.3. Result ID: 698f08d73a6c314849f51c0f.

[Reporting paper](<https://www.sota2.com/research/paper/68fa6c5a1e209131cea337d2.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa63b51e209131cea002c8.md>)

Recorded evaluation configuration: {"Backbone":"DGCNN","Training Strategy":"Meta-learning","Segmentation Module Type":"Classification-based"}. This record does not assert independent reproduction.

### Meta Learning (PN++)

mIoU: 68.5. Result ID: 698f08d73a6c314849f51c0e.

[Reporting paper](<https://www.sota2.com/research/paper/68fa6c5a1e209131cea337d2.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa63b51e209131cea002c8.md>)

Recorded evaluation configuration: {"Backbone":"PointNet++","Training Strategy":"Meta-learning","Segmentation Module Type":"Classification-based"}. This record does not assert independent reproduction.

### DGCNN

mIoU: 68. Result ID: 698f08d73a6c314849f51c0b.

[Reporting paper](<https://www.sota2.com/research/paper/68fa6c5a1e209131cea337d2.md>)

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

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

Recorded evaluation configuration: {"Backbone":"DGCNN","Segmentation Module Type":"Classification-based"}. This record does not assert independent reproduction.

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