Using the original config soft_teacher_faster_rcnn_r50_caffe_fpn_coco_180k.py with modifies w.r.t custom dataset, trained on 10% partial dataset. But the eval result is quite low, down to not even one percentile.
Is there any clue about this? Any possible mistakes made or there needs some adjustments?
2022-02-23 13:02:52,951 - mmdet.ssod - INFO - Evaluating bbox...
2022-02-23 13:02:57,158 - mmdet.ssod - INFO -
Average Precision (AP) @[ IoU=0.50:0.95 | area= all | maxDets=100 ] = 0.000
Average Precision (AP) @[ IoU=0.50 | area= all | maxDets=1000 ] = 0.002
Average Precision (AP) @[ IoU=0.75 | area= all | maxDets=1000 ] = 0.000
Average Precision (AP) @[ IoU=0.50:0.95 | area= small | maxDets=1000 ] = 0.000
Average Precision (AP) @[ IoU=0.50:0.95 | area=medium | maxDets=1000 ] = 0.001
Average Precision (AP) @[ IoU=0.50:0.95 | area= large | maxDets=1000 ] = 0.001
Average Recall (AR) @[ IoU=0.50:0.95 | area= all | maxDets=100 ] = 0.015
Average Recall (AR) @[ IoU=0.50:0.95 | area= all | maxDets=300 ] = 0.015
Average Recall (AR) @[ IoU=0.50:0.95 | area= all | maxDets=1000 ] = 0.015
Average Recall (AR) @[ IoU=0.50:0.95 | area= small | maxDets=1000 ] = 0.001
Average Recall (AR) @[ IoU=0.50:0.95 | area=medium | maxDets=1000 ] = 0.011
Average Recall (AR) @[ IoU=0.50:0.95 | area= large | maxDets=1000 ] = 0.063
Are the bounding box sizes associated with aerial images must larger or smaller than typical COCO images? If so you might want to consider adjusting the scale in the rpn_head.
Using the original config soft_teacher_faster_rcnn_r50_caffe_fpn_coco_180k.py with modifies w.r.t custom dataset, trained on 10% partial dataset. But the eval result is quite low, down to not even one percentile. Is there any clue about this? Any possible mistakes made or there needs some adjustments?
2022-02-23 13:02:52,951 - mmdet.ssod - INFO - Evaluating bbox... 2022-02-23 13:02:57,158 - mmdet.ssod - INFO - Average Precision (AP) @[ IoU=0.50:0.95 | area= all | maxDets=100 ] = 0.000 Average Precision (AP) @[ IoU=0.50 | area= all | maxDets=1000 ] = 0.002 Average Precision (AP) @[ IoU=0.75 | area= all | maxDets=1000 ] = 0.000 Average Precision (AP) @[ IoU=0.50:0.95 | area= small | maxDets=1000 ] = 0.000 Average Precision (AP) @[ IoU=0.50:0.95 | area=medium | maxDets=1000 ] = 0.001 Average Precision (AP) @[ IoU=0.50:0.95 | area= large | maxDets=1000 ] = 0.001 Average Recall (AR) @[ IoU=0.50:0.95 | area= all | maxDets=100 ] = 0.015 Average Recall (AR) @[ IoU=0.50:0.95 | area= all | maxDets=300 ] = 0.015 Average Recall (AR) @[ IoU=0.50:0.95 | area= all | maxDets=1000 ] = 0.015 Average Recall (AR) @[ IoU=0.50:0.95 | area= small | maxDets=1000 ] = 0.001 Average Recall (AR) @[ IoU=0.50:0.95 | area=medium | maxDets=1000 ] = 0.011 Average Recall (AR) @[ IoU=0.50:0.95 | area= large | maxDets=1000 ] = 0.063
Waiting for your kindly advise, thank you!