The MMDetection implementation of Mask DINO, which is taken from open-mmlab/mmdetection#9808.
The evaluation reults of the pretrained weight are presented as follows.
03/20 19:28:09 - mmengine - INFO - Panoptic Evaluation Results:
+--------+--------+--------+--------+------------+
| | PQ | SQ | RQ | categories |
+--------+--------+--------+--------+------------+
| All | 52.948 | 83.662 | 62.578 | 133 |
| Things | 58.958 | 84.918 | 69.075 | 80 |
| Stuff | 43.878 | 81.767 | 52.771 | 53 |
+--------+--------+--------+--------+------------+
03/20 19:28:20 - mmengine - INFO - Evaluating bbox...
Evaluate annotation type *bbox*
Average Precision (AP) @[ IoU=0.50:0.95 | area= all | maxDets=100 ] = 0.489
Average Precision (AP) @[ IoU=0.50 | area= all | maxDets=1000 ] = 0.686
Average Precision (AP) @[ IoU=0.75 | area= all | maxDets=1000 ] = 0.532
Average Precision (AP) @[ IoU=0.50:0.95 | area= small | maxDets=1000 ] = 0.320
Average Precision (AP) @[ IoU=0.50:0.95 | area=medium | maxDets=1000 ] = 0.520
Average Precision (AP) @[ IoU=0.50:0.95 | area= large | maxDets=1000 ] = 0.641
Average Recall (AR) @[ IoU=0.50:0.95 | area= all | maxDets=100 ] = 0.646
Average Recall (AR) @[ IoU=0.50:0.95 | area= all | maxDets=300 ] = 0.646
Average Recall (AR) @[ IoU=0.50:0.95 | area= all | maxDets=1000 ] = 0.646
Average Recall (AR) @[ IoU=0.50:0.95 | area= small | maxDets=1000 ] = 0.464
Average Recall (AR) @[ IoU=0.50:0.95 | area=medium | maxDets=1000 ] = 0.680
Average Recall (AR) @[ IoU=0.50:0.95 | area= large | maxDets=1000 ] = 0.811
03/20 19:29:00 - mmengine - INFO - bbox_mAP_copypaste: 0.489 0.686 0.532 0.320 0.520 0.641
03/20 19:29:00 - mmengine - INFO - Evaluating segm...
Evaluate annotation type *segm*
Average Precision (AP) @[ IoU=0.50:0.95 | area= all | maxDets=100 ] = 0.443
Average Precision (AP) @[ IoU=0.50 | area= all | maxDets=1000 ] = 0.671
Average Precision (AP) @[ IoU=0.75 | area= all | maxDets=1000 ] = 0.482
Average Precision (AP) @[ IoU=0.50:0.95 | area= small | maxDets=1000 ] = 0.244
Average Precision (AP) @[ IoU=0.50:0.95 | area=medium | maxDets=1000 ] = 0.477
Average Precision (AP) @[ IoU=0.50:0.95 | area= large | maxDets=1000 ] = 0.636
Average Recall (AR) @[ IoU=0.50:0.95 | area= all | maxDets=100 ] = 0.583
Average Recall (AR) @[ IoU=0.50:0.95 | area= all | maxDets=300 ] = 0.583
Average Recall (AR) @[ IoU=0.50:0.95 | area= all | maxDets=1000 ] = 0.583
Average Recall (AR) @[ IoU=0.50:0.95 | area= small | maxDets=1000 ] = 0.393
Average Recall (AR) @[ IoU=0.50:0.95 | area=medium | maxDets=1000 ] = 0.626
Average Recall (AR) @[ IoU=0.50:0.95 | area= large | maxDets=1000 ] = 0.762
03/20 19:29:47 - mmengine - INFO - segm_mAP_copypaste: 0.443 0.671 0.482 0.244 0.477 0.636
03/20 19:29:48 - mmengine - INFO - Iter(test) [5000/5000] coco_panoptic/PQ: 52.9484 coco_panoptic/SQ: 83.6622 coco_panoptic/RQ: 62.5779 coco_panoptic/PQ_th: 58.9575 coco_panoptic/SQ_th: 84.9180 coco_panoptic/RQ_th: 69.0747 coco_panoptic/PQ_st: 43.8780 coco_panoptic/SQ_st: 81.7665 coco_panoptic/RQ_st: 52.7714 coco/bbox_mAP: 0.4890 coco/bbox_mAP_50: 0.6860 coco/bbox_mAP_75: 0.5320 coco/bbox_mAP_s: 0.3200 coco/bbox_mAP_m: 0.5200 coco/bbox_mAP_l: 0.6410 coco/segm_mAP: 0.4430 coco/segm_mAP_50: 0.6710 coco/segm_mAP_75: 0.4820 coco/segm_mAP_s: 0.2440 coco/segm_mAP_m: 0.4770 coco/segm_mAP_l: 0.6360