Caffe is one of the most popular deep learning frameworks. It is written in C++.
From the official site:
"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 by community contributors."
The library is primarily targeted at Linux and macOS X but unofficial Android, iOS, and Windows ports are also compatible. Caffe supports CUDA acceleration but can also be executed on the CPU alone. On iOS, it uses only CPU. The interfaces include C++, command line, Python, and MATLAB. Caffe provides recurrent and convolutional neural networks. To define a network, you need to describe its structure in a config file in a special format.
ModelZoo ...
