rpn_probs and rpn_class_logits foreground vs background #870
Comments
|
In rpn_class_loss_graph, BG/FG's label are -1/+1 and then 0/1. So the comment in rpn_class_loss_graph should be a typo. |
keineahnung2345
added a commit
to keineahnung2345/Mask_RCNN
that referenced
this issue
Jan 10, 2019
waleedka
pushed a commit
that referenced
this issue
Jan 28, 2019
luccauchon
added a commit
to luccauchon/Mask_RCNN
that referenced
this issue
Feb 4, 2019
* Fix visualize activations (matterport#1211) Fix visualize activations (Squashed 4 commits by @keineahnung2345) # Get activations of a few sample layers activations = model.run_graph([image], [ ("input_image", model.keras_model.get_layer("input_image").output) ]) leads to the error: InvalidArgumentError: input_image:0 is both fed and fetched. Revise the code according to https://stackoverflow.com/questions/39307108/placeholder-20-is-both-fed-and-fetched * Fix typo in config.py * Fix typo in rpn_class_loss_graph Fix matterport#870 * Reversed order (1) and (2) in installation Reference to matterport#1255 * remove unused code (matterport#1227) Remove unused code
Closed
aneeshchauhan
pushed a commit
to aneeshchauhan/Mask_RCNN
that referenced
this issue
Jul 9, 2019
smsegal
pushed a commit
to smsegal/mask_rcnn
that referenced
this issue
Nov 18, 2020
withyou53
pushed a commit
to withyou53/mask_r-cnn_for_object_detection
that referenced
this issue
Sep 27, 2023
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
A quick question about the rpn_probs and rpn_class_logits. In the rpn_graph function when rpn_probs is first defined it is stated that the last dimension is separated into BG/FG. However, when using rpn_class_logits in the rpn_class_loss_graph function, it's last dimension is defined to be FG/BG, which is the reverse of rpn_probs. Is there a reason why it is swapped?
The text was updated successfully, but these errors were encountered: