# 3D Object Detection on KITTI (test) (AP R40)

Source: [3D Object Detection on KITTI (test) (AP R40)](<https://www.sota2.com/research/sota/3d-object-detection-on-kitti-test-ap-r40>)

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

Data updated: 2026-06-02T08:56:09.101Z

Evaluation of 3D Object Detection performance on the KITTI test set, reporting Average Precision based on the AP R40 metric across Car, Pedestrian, and Cyclist categories, broken down by Easy, Moderate, and Hard difficulty levels.

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 Detection |
| Dataset (including any stated split) | KITTI (test) |
| Metric | Car Easy AP R40 |
| Unit | PERCENT |
| Direction | Higher is better |
| Results loaded | 4 of 4 |
| 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

- [Car Easy AP R40](<https://www.sota2.com/research/sota/3d-object-detection-on-kitti-test-ap-r40.md?metric=Car%20Easy%20AP%20R40>): Higher is better; unit: PERCENT.

- [Car Moderate AP R40](<https://www.sota2.com/research/sota/3d-object-detection-on-kitti-test-ap-r40.md?metric=Car%20Moderate%20AP%20R40>): Higher is better; unit: PERCENT.

- [Car Hard AP R40](<https://www.sota2.com/research/sota/3d-object-detection-on-kitti-test-ap-r40.md?metric=Car%20Hard%20AP%20R40>): Higher is better; unit: PERCENT.

- [Pedestrian Easy AP R40](<https://www.sota2.com/research/sota/3d-object-detection-on-kitti-test-ap-r40.md?metric=Pedestrian%20Easy%20AP%20R40>): Higher is better; unit: PERCENT.

- [Pedestrian Moderate AP R40](<https://www.sota2.com/research/sota/3d-object-detection-on-kitti-test-ap-r40.md?metric=Pedestrian%20Moderate%20AP%20R40>): Higher is better; unit: PERCENT.

- [Pedestrian Hard AP R40](<https://www.sota2.com/research/sota/3d-object-detection-on-kitti-test-ap-r40.md?metric=Pedestrian%20Hard%20AP%20R40>): Higher is better; unit: PERCENT.

- [Cyclist Easy AP R40](<https://www.sota2.com/research/sota/3d-object-detection-on-kitti-test-ap-r40.md?metric=Cyclist%20Easy%20AP%20R40>): Higher is better; unit: PERCENT.

- [Cyclist Moderate AP R40](<https://www.sota2.com/research/sota/3d-object-detection-on-kitti-test-ap-r40.md?metric=Cyclist%20Moderate%20AP%20R40>): Higher is better; unit: PERCENT.

- [Cyclist Hard AP R40](<https://www.sota2.com/research/sota/3d-object-detection-on-kitti-test-ap-r40.md?metric=Cyclist%20Hard%20AP%20R40>): Higher is better; unit: PERCENT.

- [Average Easy AP R40](<https://www.sota2.com/research/sota/3d-object-detection-on-kitti-test-ap-r40.md?metric=Average%20Easy%20AP%20R40>): Higher is better; unit: PERCENT.

- [Average Moderate AP R40](<https://www.sota2.com/research/sota/3d-object-detection-on-kitti-test-ap-r40.md?metric=Average%20Moderate%20AP%20R40>): Higher is better; unit: PERCENT.

- [Average Hard AP R40](<https://www.sota2.com/research/sota/3d-object-detection-on-kitti-test-ap-r40.md?metric=Average%20Hard%20AP%20R40>): Higher is better; unit: PERCENT.

## Recorded results

| Method | Car Easy AP R40 | Result date | Report paper |
| --- | --- | --- | --- |
| FSDV2 | 86.56 | Not recorded | 6a1e63cadbd7150506f853fa |
| FSDV2 + SPPCC | 86.53 | Not recorded | 6a1e63cadbd7150506f853fa |
| VoxelNext | 83.88 | Not recorded | 6a1e63cadbd7150506f853fa |
| VoxelNext + SPPCC | 82.45 | Not recorded | 6a1e63cadbd7150506f853fa |

### FSDV2

Car Easy AP R40: 86.56. Result ID: 6a1e9a94fbc373592c18e2ad.

[Reporting paper](<https://www.sota2.com/research/paper/6a1e63cadbd7150506f853fa.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa70481e209131cea4a2c0.md>)

[Implementation](<https://github.com/tusen-ai/SST>)

Recorded evaluation configuration: {"Base detector":"FSDV2"}. This record does not assert independent reproduction.

### FSDV2 + SPPCC

Car Easy AP R40: 86.53. Result ID: 6a1e9a94fbc373592c18e2ae.

[Reporting paper](<https://www.sota2.com/research/paper/6a1e63cadbd7150506f853fa.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6a1e63cadbd7150506f853fa.md>)

Recorded evaluation configuration: {"Base detector":"FSDV2","Module":"SPPCC"}. This record does not assert independent reproduction.

### VoxelNext

Car Easy AP R40: 83.88. Result ID: 6a1e9a94fbc373592c18e2af.

[Reporting paper](<https://www.sota2.com/research/paper/6a1e63cadbd7150506f853fa.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6ed21e209131cea42682.md>)

[Implementation](<https://www.sota2.com/github.com/dvlab-research/VoxelNeXt>)

Recorded evaluation configuration: {"Base detector":"VoxelNext"}. This record does not assert independent reproduction.

### VoxelNext + SPPCC

Car Easy AP R40: 82.45. Result ID: 6a1e9a94fbc373592c18e2b0.

[Reporting paper](<https://www.sota2.com/research/paper/6a1e63cadbd7150506f853fa.md>)

[Method origin paper](<https://www.sota2.com/research/paper/6a1e63cadbd7150506f853fa.md>)

Recorded evaluation configuration: {"Base detector":"VoxelNext","Module":"SPPCC"}. This record does not assert independent reproduction.

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