CharlesShang / TFFRCNN

FastER RCNN built on tensorflow
MIT License
874 stars 418 forks source link

AssertionError in adding bounding box targets #66

Open hadign20 opened 7 years ago

hadign20 commented 7 years ago

I'm trying to run faster rcnn for a new dataset. I have made some changes in pascal_voc.py based on the new dataset, but when I run the training, I get this error:

Computing bounding-box regression targets...
Traceback (most recent call last):
  File "./tools/train_net.py", line 96, in <module>
    max_iters=args.max_iters)
  File "/home/hadi/Software/tensorflow/Faster-RCNN_TF/tools/../lib/fast_rcnn/train.py", line 263, in train_net
    sw = SolverWrapper(sess, saver, network, imdb, roidb, output_dir, pretrained_model=pretrained_model)
  File "/home/hadi/Software/tensorflow/Faster-RCNN_TF/tools/../lib/fast_rcnn/train.py", line 39, in __init__
    self.bbox_means, self.bbox_stds = rdl_roidb.add_bbox_regression_targets(roidb)
  File "/home/hadi/Software/tensorflow/Faster-RCNN_TF/tools/../lib/roi_data_layer/roidb.py", line 48, in add_bbox_regression_targets
    assert len(roidb) > 0
AssertionError

How can I fix this error?