Skip to content

Custom layers for deep learning networks#11129

Merged
vpisarev merged 2 commits intoopencv:3.4from
dkurt:dnn_custom_layers
Apr 24, 2018
Merged

Custom layers for deep learning networks#11129
vpisarev merged 2 commits intoopencv:3.4from
dkurt:dnn_custom_layers

Conversation

@dkurt
Copy link
Copy Markdown
Member

@dkurt dkurt commented Mar 21, 2018

This pullrequest changes

  • Custom layers from Caffe and TensorFlow.
  • Fix some internal code typos.
  • Unite duplicated OpenCL and CPU layers tests.

resolves: #10985
resolves: #10888
closes: #11050 (ResizeBilinear)
closes: #10889 (SELU activation)

Merge with extra: opencv/opencv_extra#448

@dkurt dkurt force-pushed the dnn_custom_layers branch 2 times, most recently from 8e18376 to e3e465b Compare March 21, 2018 10:28
@dkurt dkurt force-pushed the dnn_custom_layers branch from e3e465b to ddf6036 Compare March 25, 2018 09:52
@dkurt dkurt changed the title Custom layers for networks from Caffe and TensorFlow Custom layers for deep learning networks Mar 25, 2018
the following methods:

~~~~~~~~~~~~~{.cpp}
class MyLayer : public cv::dnn::Layer
Copy link
Copy Markdown
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

I believe it is better to place whole sample code into a separate file (which could be checked by compiler) and put here snippets only.
See this example.

Copy link
Copy Markdown
Member Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

@alalek, thanks! Done.

@dkurt dkurt force-pushed the dnn_custom_layers branch 2 times, most recently from 3e04dc9 to 1513b01 Compare March 28, 2018 09:44
@dkurt dkurt force-pushed the dnn_custom_layers branch 2 times, most recently from a240a9d to 38fe601 Compare April 2, 2018 07:50
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Projects

None yet

Development

Successfully merging this pull request may close these issues.

3 participants