search

endernewton / tf-faster-rcnn

Tensorflow Faster RCNN for Object Detection
https://arxiv.org/pdf/1702.02138.pdf
MIT License
3.65k stars 1.58k forks source link

bug in resnet_v1._crop_pool_layer? #366

Open akonsk opened 6 years ago

akonsk commented 6 years ago

hi, thanks for sharing this great work!

https://github.com/endernewton/tf-faster-rcnn/blob/0e61bacf863f9f466b54770f35a130514e85cac6/lib/nets/resnet_v1.py#L60

i think you meant these lines to be: height = tf.to_float(bottom_shape[1]) np.float32(self._feat_stride[0]) - 1. width = tf.to_float(bottom_shape[2]) np.float32(self._feat_stride[0]) - 1.

or do i miss something?

thanks!

sun-su commented 5 years ago

Same question here. Anyone can explain this please?