Caffe

Caffe is a deep learning framework made with expression, speed, and modularity in mind. It is developed by Berkeley AI Research (BAIR)/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
• BAIR reference models and the community model zoo
• Installation instructions

and step-by-step examples.

Custom distributions

• Intel Caffe (Optimized for CPU and support for multi-node), in particular Intel® Xeon processors.
• OpenCL Caffe e.g. for AMD or Intel devices.
• Windows Caffe

Community

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!

License and Citation

Caffe is released under the BSD 2-Clause license. The BAIR/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}
}

Added By Myself

How to build Caffe on Ubuntu 16.04

If you just want to use the officially-compiled caffe bins instead of modifying caffe code, using a docker is a better option. Currently there are two very-well maintained caffe docker.

1. bvlc/caffe:cpu
2. ufoym/deepo:cpu

Detailed description of how to use bvlc/caffe:cpu is located Here.

Steps to build caffe on a Ubuntu 16.04 machine.

Prerequisites: Install the dependent packages

sudo apt-get install libprotobuf-dev libleveldb-dev libsnappy-dev libopencv-dev libhdf5-serial-dev protobuf-compiler
sudo apt-get install --no-install-recommends libboost-all-dev
sudo apt-get install libatlas-base-dev
sudo apt-get install libgflags-dev libgoogle-glog-dev liblmdb-dev
for req in $(cat ./python/requirements.txt); do pip install $req; done

Fix 3 installation errors due to using Python 2.x version. Matplotlib error was ignored.

pip install 'scikit-image<0.15'
pip install 'networkx==2.2'
pip install IPython==5.0 --user

As you will see when you run the classification of an image, an skimage.io will occur as flowing. ImportError: No module named skimage.io Run the following command to install skimage.

sudo apt-get install python-skimage

For reference, if you meet new errors regarding of python packages, you can use the following command instead of using pip install

sudo apt-get install python-matplotlib python-numpy python-pil python-scipy
sudo apt-get install build-essential cython

Prepare the Makefile.config

cp Makefile.config.example Makfile.config
Uncommend CPU_ONLY := 1 in Makefile.config and save the file

Build the project

cd build
make clean
cmake ..
make all

Install the bins

make install

Test the installation

make runtest

If you follow the make method on BVLC/caffe you will run into two errors, so this way is not recommended. If you insist to use make, fixes are listed below.

Error 1: hdf5.h missing issue

Error 2: libcaffe.so.1.0.0 error issue

A classification example

Before running classification, we need to download the models first. Run the following command to download the model file.

scripts/download_model_binary.py models/bvlc_reference_caffenet

Classification command example

python python/classify.py examples/images/cat.jpg foo

Note that the results are saved into foo.npy, which is a numpy data file. Use the following command to see the results in python.

import numpy as np
np.load('foo.npy')