mirror/opencv
1
0
Fork 0
mirror of https://github.com/opencv/opencv.git synced 2024-11-27 20:50:25 +08:00
Code Issues Actions Packages Projects Releases Wiki Activity

Updated GSoC_2018 (markdown)

Vadim Pisarevsky 2018-01-15 21:05:51 +03:00
parent 13d7f9e15b
commit f7f18ca8f3
1 changed files with 1 additions and 0 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

1
GSoC_2018.md

@ -182,6 +182,7 @@ Improving [DNN module](https://github.com/opencv/opencv_contrib/tree/master/modu
- Advanced visualizing of deep learning models (with input-output blobs dimensions, layer types, connections and so on). It can be either based on third-party libraries solution or built on OpenCV's drawing functions own implementation.
- Implementing various network compression algorithms, such as quantization, factorization-based compression etc. That will likely involve both scripts for TF/Caffe and the corresponding modifications in DNN to support such compressed models.
- Enable supporting of the most popular deep learning architectures and classes of architectures that DNN does not support yet: DenseNet, GAN's etc. Implement missed layers, check output accuracy, write samples for online-available models.
- Implement ONNX format parser. We already support Caffe 1 (with various extensions), TF, Torch and part of Darknet. Adding yet another parser is time-consuming but quite straight-forward thing.
### 2b. Curating and optimizing valuable published deep neural networks and turning them into compact models.