2D Class-agnostic Object Detection on Waymo (train)
4,320AP50ClusterNet
Evaluation Results
| Method | Links | ||||
|---|---|---|---|---|---|
| ClusterNetannotation setting=4D data, #images=158k, #bboxes=02022.10 | 4,320 | 1,840 | 5,650 | 8,180 | |
| FuseMODNetannotation setting=zero-shot video, #images=158k, #bboxes=0, pre-trained with manual annotations=true2022.10 | 2,490 | 370 | 2,960 | 5,810 | |
| RVOSannotation setting=zero-shot video, #images=158k, #bboxes=0, pre-trained with manual annotations=true2022.10 | 1,430 | 0 | 660 | 5,410 | |
| COSNetannotation setting=zero-shot video, #images=158k, #bboxes=0, pre-trained with manual annotations=true2022.10 | 1,030 | 0 | 250 | 4,540 | |
| MCGannotation setting=still-image, #images=158k, #bboxes=02022.10 | 690 | 0 | 300 | 3,750 | |
| LOSTannotation setting=still-image, #images=158k, #bboxes=02022.10 | 190 | 0 | 100 | 760 | |
| Felzenszwalb Segmentationannotation setting=still-image, #images=158k, #bboxes=02022.10 | 40 | 0 | 50 | 110 |