search

jwyang / faster-rcnn.pytorch

A faster pytorch implementation of faster r-cnn
MIT License
7.7k stars 2.33k forks source link

Testing: evaluate_detections "NotImplementedError" #488

Open mahaling opened 5 years ago

mahaling commented 5 years ago

I got a vgg16 successfully trained on a custom dataset and it loaded perfectly while testing. The test_net.py code has a "imdb.evaluate_detections" at the end and this function is not implemented yet. I am getting the following error. Traceback (most recent call last): File "test_net.py", line 333, in imdb.evaluate_detections(all_boxes, output_dir) File "/faster_rcnn/lib/datasets/imdb.py", line 108, in evaluate_detections raise NotImplementedError NotImplementedError

Is there something I am missing here?. Any help would be appreciated. Thanks.

berniel commented 5 years ago

Same question here. Besides, should I just uncomment the evaluate_recall function if I want to evaluate recall?