# Object Detection on SODA10M (test)

Source: [Object Detection on SODA10M (test)](<https://www.sota2.com/research/sota/object-detection-on-soda10m-test>)

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

Data updated: 2026-02-26T05:23:25.756Z

Evaluation of standard Object Detection performance on the test split of the SODA10M dataset, reporting class-specific Average Precision metrics (e.g., AP@50).

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 | Object Detection |
| Dataset (including any stated split) | SODA10M (test) |
| Metric | AP@50 (Car) |
| Unit | PERCENT |
| Direction | Higher is better |
| Results loaded | 3 of 3 |
| 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

- [AP@50 (Car)](<https://www.sota2.com/research/sota/object-detection-on-soda10m-test.md?metric=AP%4050%20(Car)>): Higher is better; unit: PERCENT.

- [AP@50 (Pedestrian)](<https://www.sota2.com/research/sota/object-detection-on-soda10m-test.md?metric=AP%4050%20(Pedestrian)>): Higher is better; unit: PERCENT.

- [AP@50 (Cyclist)](<https://www.sota2.com/research/sota/object-detection-on-soda10m-test.md?metric=AP%4050%20(Cyclist)>): Higher is better; unit: PERCENT.

## Recorded results

| Method | AP@50 (Car) | Result date | Report paper |
| --- | --- | --- | --- |
| Fully Supervised | 84.8 | Not recorded | 68fa6c2a1e209131cea3277e |
| Dynamic Unsupervised Adaptation | 49.4 | Not recorded | 68fa6c2a1e209131cea3277e |
| Source-only | 43.6 | Not recorded | 68fa6c2a1e209131cea3277e |

### Fully Supervised

AP@50 (Car): 84.8. Result ID: 698f50673a6c314849f7f7d0.

[Reporting paper](<https://www.sota2.com/research/paper/68fa6c2a1e209131cea3277e.md>)

Recorded evaluation configuration: {"Architecture":"YOLOv3","Training Split":"KITTI train split","Supervision Protocol":"Fully Supervised (FS)"}. This record does not assert independent reproduction.

### Dynamic Unsupervised Adaptation

AP@50 (Car): 49.4. Result ID: 698f50673a6c314849f7f7cf.

[Reporting paper](<https://www.sota2.com/research/paper/68fa6c2a1e209131cea3277e.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6c2a1e209131cea3277e.md>)

[Implementation](<https://github.com/jmiemirza/DUA>)

Recorded evaluation configuration: {"Architecture":"YOLOv3","Training Split":"KITTI train split","Supervision Protocol":"Dynamic Unsupervised Adaptation (DUA)"}. This record does not assert independent reproduction.

### Source-only

AP@50 (Car): 43.6. Result ID: 698f50673a6c314849f7f7ce.

[Reporting paper](<https://www.sota2.com/research/paper/68fa6c2a1e209131cea3277e.md>)

Recorded evaluation configuration: {"Architecture":"YOLOv3","Training Split":"KITTI train split","Supervision Protocol":"Source-only (SO)"}. This record does not assert independent reproduction.

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