search

facebookresearch / Detectron

FAIR's research platform for object detection research, implementing popular algorithms like Mask R-CNN and RetinaNet.
Apache License 2.0
26.26k stars 5.45k forks source link

Average Recall (AR) @[ IoU=0.50:0.95 | area= large | maxDets=100 ] = -1.000 #588

Open travis0925 opened 6 years ago

travis0925 commented 6 years ago

I trained the model with my custom dataset, 2 classes(including background), the config file is retinanet_R-50-FPN_1X.yaml. And I got the test result like this: INFO json_dataset_evaluator.py: 222: ~~ Mean and per-category AP @ IoU=[0.50,0.95] ~~ INFO json_dataset_evaluator.py: 223: 30.4 INFO json_dataset_evaluator.py: 231: 30.4 INFO json_dataset_evaluator.py: 232: ~~ Summary metrics ~~ Average Precision (AP) @[ IoU=0.50:0.95 | area= all | maxDets=100 ] = 0.304 Average Precision (AP) @[ IoU=0.50 | area= all | maxDets=100 ] = 0.804 Average Precision (AP) @[ IoU=0.75 | area= all | maxDets=100 ] = 0.100 Average Precision (AP) @[ IoU=0.50:0.95 | area= small | maxDets=100 ] = 0.319 Average Precision (AP) @[ IoU=0.50:0.95 | area=medium | maxDets=100 ] = 0.286 Average Precision (AP) @[ IoU=0.50:0.95 | area= large | maxDets=100 ] = -1.000 Average Recall (AR) @[ IoU=0.50:0.95 | area= all | maxDets= 1 ] = 0.377 Average Recall (AR) @[ IoU=0.50:0.95 | area= all | maxDets= 10 ] = 0.492 Average Recall (AR) @[ IoU=0.50:0.95 | area= all | maxDets=100 ] = 0.497 Average Recall (AR) @[ IoU=0.50:0.95 | area= small | maxDets=100 ] = 0.510 Average Recall (AR) @[ IoU=0.50:0.95 | area=medium | maxDets=100 ] = 0.468 Average Recall (AR) @[ IoU=0.50:0.95 | area= large | maxDets=100 ] = -1.000

You guys have any idea of that why I got the AP of value -1 when area=large? This doesn't happen every time, I try to figure out the theory behind this, can you help me solve this problem if you know anything about this?

AlexLiuJS commented 6 years ago

hey guys, have you defeated this problem?(AP -1) i got a similar trouble but more bad :( my problem is AP -1, AP50 -1, AP75 -1, APs -1, APm -1, APl -1 have you any idea with that? thanks a lot!!!