Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Add cross entropy layer for pixel classification #2876

Closed

Conversation

wanghaoshuang
Copy link
Contributor

  1. Add pixel_cross_entropy_cost layer
  2. Add INPUT_MULTI_LABEL type for generating pixel label data in LayerGradUtil.cpp
    fix Add cross entropy layer for pixel classification #2875

1. Add switch function for switching image dimensions order
2. Add CpuMatrix::backwardSoftmax function
3. Add pixel softmax layer, python wrapper and grad_test
1. Add pixel_cross_entropy_cost layer
2. Add INPUT_MULTI_LABEL type for generating pixel label data in LayerGradUtil.cpp
@qingqing01
Copy link
Contributor

@wanghaoshuang as we discussed earlier, I still think there is no need to develop pixel_cross_entropy_cost. And this layer is not a general operation and a litter strange. We can expose the switch_layer for users and use the switch_layer, resize_layer, softmax and cross_entropy_cost for the pixel classification:

switch_layer -> resize_layer -> softmax -> cross_entropy_cost

@wanghaoshuang wanghaoshuang deleted the pixel_cross_entropy branch July 19, 2017 06:56
@wanghaoshuang
Copy link
Contributor Author

@qingqing01 I implement switch_layer with reshape option in PR#2788 which support for softmax activation.
So i closed this PR.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

Successfully merging this pull request may close these issues.

Add cross entropy layer for pixel classification
2 participants