Here are some of the popular deep learning libraries used in research and commercial applications:
Figure 1: Popular deep learning libraries
- TensorFlow: This is an open source software library for numerical computation using data flow graphs. The TensorFlow library is designed for machine intelligence and developed by the Google Brain team. The main aim of this library is to perform machine learning and deep neural network research. It can be used in a wide variety of other domains as well (https://www.tensorflow.org/).
- Theano: This is an open source Python library (http://deeplearning.net/software/theano/) that enables us to optimize and evaluate mathematical expressions involving multidimensional arrays efficiently. Theano is primarily developed by the machine learning group at the University of Montreal , Canada.
- Torch: Torch is again a scientific computing framework with wide support for machine learning algorithms that puts GPUs first. It's very efficient, being built on the scripting language LuaJIT and has an underlying C/CUDA implementation (http://torch.ch/).
- Caffe: Caffe (http://caffe.berkeleyvision.org/) is a deep learning library made with a focus on modularity, speed, and expression. It is developed by the Berkeley Vision and Learning Centre (BVLC).