Caffe is a deep learning framework made with expression, speed, and modularity in mind. It is developed by the Berkeley Vision and Learning Center (BVLC) and community contributors.
Check out the project site for all the details like
- DIY Deep Learning for Vision with Caffe
- Tutorial Documentation
- BVLC reference models and the community model zoo
- Installation instructions
and step-by-step examples.
Please join the caffe-users group or gitter chat to ask questions and talk about methods and models. Framework development discussions and thorough bug reports are collected on Issues.
Happy brewing!
Caffe is released under the BSD 2-Clause license. The BVLC reference models are released for unrestricted use.
Please cite Caffe in your publications if it helps your research:
@article{jia2014caffe,
Author = {Jia, Yangqing and Shelhamer, Evan and Donahue, Jeff and Karayev, Sergey and Long, Jonathan and Girshick, Ross and Guadarrama, Sergio and Darrell, Trevor},
Journal = {arXiv preprint arXiv:1408.5093},
Title = {Caffe: Convolutional Architecture for Fast Feature Embedding},
Year = {2014}
}
There exists three layers related to our landmark project: yolo_loss_layer, perceptual_loss_layer and patch_semantic_layer. I finished the first two layers at present (source codes for patch_semantic_layer is also finished), both source codes under 'src/caffe/layers' directory and test codes (test whether Forward and Backward are correct) under 'src/caffe/test' directory. And the only one header file is 'deep_landmark_layers.hpp' under 'include/caffe' directory.
The reference paper for yolo_loss_layer is: You Only Look Once: Unified, Real-Time Object Detection The reference paper for perceptual_loss_layer is: UNSUPERVISED CONVOLUTIONAL NEURAL NETWORKS FOR MOTION ESTIMATION and Determining optical flow. For the paper Determining optical flow, you just need to figure out the computation of the three partial derivatives (this is analogously what patch_semantic_layer does). If you got any problems, feel free to let me know.