BDD
Benchmarks
| Task Name | Dataset Name | SOTA Result | Trend | |
|---|---|---|---|---|
| Semantic Segmentation | BDD-100K (val) | mIoU66.28 | 102 | |
| Semantic Segmentation | BDD100K | mIoU93.4 | 78 | |
| Semantic Segmentation | BDD (val) | mIoU63.6 | 47 | |
| Drivable Area Segmentation | BDD100K v1.0 (test) | mIoU (%)93.2 | 41 | |
| Semantic Segmentation | BDD Night (test) | Mean IoU43.14 | 31 | |
| Semantic Segmentation | BDD target domain | mIoU41.73 | 20 | |
| Semantic Segmentation | BDD | mIoU68.08 | 19 | |
| Segmentation | BDD100K | Failure Rate0 | 12 | |
| Scene Graph to Image Generation | BDD100k | FID46.1 | 12 | |
| Object Detection | BDD | mAP5033.4 | 8 | |
| Semantic Segmentation | BDD100K BDD-19 | mIoU61.8 | 7 | |
| Segmentation | BDD-10K Real-World Degradations (unseen) | IoU90.3 | 7 | |
| Object Detection | BDD100k (unlabeled target set) | Truck Count53.4 | 7 | |
| Open-vocabulary semantic segmentation | BDD100K | mIoU60.06 | 6 | |
| Semantic Segmentation | BDD-40 (val) | mIoU47.4 | 6 | |
| Panoptic Segmentation | BDD-40 (val) | PQ19.4 | 6 | |
| Multiple Object Tracking | BDD (test) | mIDF160 | 5 | |
| Semantic Segmentation | BDD (test) | mIoU35.2 | 5 | |
| Semantic Segmentation | BDD150 | mPA (%)52.96 | 5 | |
| Image Deraining | BDD150 | PSNR22.48 | 5 | |
| Object Detection | BDD350 | Precision41.04 | 5 | |
| Image Deraining | BDD350 | PSNR17.85 | 5 | |
| Panoptic Segmentation | BDD | PQ33 | 4 | |
| Scene Graph to Image Generation | BDD (randomly sampled) | Rank 1 Accuracy (%)0.3472 | 4 | |
| Pristine-Degraded Separability | BDD heavy rain | AUROC88.52 | 2 |
Showing 25 of 28 rows