# Object Detection on BDD100K (500 tracked image pairs)

Source: [Object Detection on BDD100K (500 tracked image pairs)](<https://www.sota2.com/research/sota/object-detection-on-bdd100k-500-tracked-image-pairs>)

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

Data updated: 2026-07-08T06:57:19.797Z

Evaluation of Object Detection performance specifically on the 500 tracked image pairs subset of the BDD100K dataset, reporting fundamental detection statistics like True Positives, False Positives, False Negatives, Average Precision, and Average Recall.

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) | BDD100K (500 tracked image pairs) |
| Metric | True Positives (TP) |
| Unit | SCALAR |
| 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

- [True Positives (TP)](<https://www.sota2.com/research/sota/object-detection-on-bdd100k-500-tracked-image-pairs.md?metric=True%20Positives%20(TP)>): Higher is better; unit: SCALAR.

- [False Positives (FP)](<https://www.sota2.com/research/sota/object-detection-on-bdd100k-500-tracked-image-pairs.md?metric=False%20Positives%20(FP)>): Lower is better; unit: SCALAR.

- [False Negatives (FN)](<https://www.sota2.com/research/sota/object-detection-on-bdd100k-500-tracked-image-pairs.md?metric=False%20Negatives%20(FN)>): Lower is better; unit: SCALAR.

- [Average Precision (AP)](<https://www.sota2.com/research/sota/object-detection-on-bdd100k-500-tracked-image-pairs.md?metric=Average%20Precision%20(AP)>): Higher is better; unit: PERCENT.

- [Average Recall (AR)](<https://www.sota2.com/research/sota/object-detection-on-bdd100k-500-tracked-image-pairs.md?metric=Average%20Recall%20(AR)>): Higher is better; unit: PERCENT.

## Recorded results

| Method | True Positives (TP) | Result date | Report paper |
| --- | --- | --- | --- |
| Pixel Replace | 2223 | Not recorded | 6a4dd9af4da6617fe9198c08 |
| LaMa Inpaint | 1528 | Not recorded | 6a4dd9af4da6617fe9198c08 |
| Telea Inpaint | 1413 | Not recorded | 6a4dd9af4da6617fe9198c08 |
| Navier-Stokes | 1408 | Not recorded | 6a4dd9af4da6617fe9198c08 |
| Unrecovered VOA | 1346 | Not recorded | 6a4dd9af4da6617fe9198c08 |
| Stable Diffusion | 603 | Not recorded | 6a4dd9af4da6617fe9198c08 |
| Median Filter | 488 | Not recorded | 6a4dd9af4da6617fe9198c08 |

### Pixel Replace

True Positives (TP): 2223. Result ID: 6a4df4c460b88bb339d6a11c.

[Reporting paper](<https://www.sota2.com/research/paper/6a4dd9af4da6617fe9198c08.md>)

Recorded evaluation configuration: {"Detection Model":"YOLOv8l","IoU Threshold":"\>= 0.50","Sample Size (n)":"500"}. This record does not assert independent reproduction.

### LaMa Inpaint

True Positives (TP): 1528. Result ID: 6a4df4c460b88bb339d6a11d.

[Reporting paper](<https://www.sota2.com/research/paper/6a4dd9af4da6617fe9198c08.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6b311e209131cea2cb89.md>)

[Implementation](<https://github.com/saic-mdal/lama>)

Recorded evaluation configuration: {"Detection Model":"YOLOv8l","IoU Threshold":"\>= 0.50","Sample Size (n)":"500"}. This record does not assert independent reproduction.

### Telea Inpaint

True Positives (TP): 1413. Result ID: 6a4df4c460b88bb339d6a11e.

[Reporting paper](<https://www.sota2.com/research/paper/6a4dd9af4da6617fe9198c08.md>)

Recorded evaluation configuration: {"Detection Model":"YOLOv8l","IoU Threshold":"\>= 0.50","Sample Size (n)":"500"}. This record does not assert independent reproduction.

### Navier-Stokes

True Positives (TP): 1408. Result ID: 6a4df4c460b88bb339d6a11f.

[Reporting paper](<https://www.sota2.com/research/paper/6a4dd9af4da6617fe9198c08.md>)

Recorded evaluation configuration: {"Detection Model":"YOLOv8l","IoU Threshold":"\>= 0.50","Sample Size (n)":"500"}. This record does not assert independent reproduction.

### Unrecovered VOA

True Positives (TP): 1346. Result ID: 6a4df4c460b88bb339d6a11b.

[Reporting paper](<https://www.sota2.com/research/paper/6a4dd9af4da6617fe9198c08.md>)

Recorded evaluation configuration: {"Detection Model":"YOLOv8l","IoU Threshold":"\>= 0.50","Sample Size (n)":"500"}. This record does not assert independent reproduction.

### Stable Diffusion

True Positives (TP): 603. Result ID: 6a4df4c460b88bb339d6a121.

[Reporting paper](<https://www.sota2.com/research/paper/6a4dd9af4da6617fe9198c08.md>)

[Method origin paper](<https://www.sota2.com/research/paper/68fa6c391e209131cea32e4a.md>)

[Implementation](<https://github.com/CompVis/latent-diffusion>)

Recorded evaluation configuration: {"Detection Model":"YOLOv8l","IoU Threshold":"\>= 0.50","Sample Size (n)":"400","VOA Scope":"structured VOAs only"}. This record does not assert independent reproduction.

### Median Filter

True Positives (TP): 488. Result ID: 6a4df4c460b88bb339d6a120.

[Reporting paper](<https://www.sota2.com/research/paper/6a4dd9af4da6617fe9198c08.md>)

Recorded evaluation configuration: {"Detection Model":"YOLOv8l","IoU Threshold":"\>= 0.50","Sample Size (n)":"500"}. This record does not assert independent reproduction.

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