There is an enormous wealth of open source deep learning tools coming from the computer vision space. Among these Keras, Caffe, Torch, TensorFlow, Theano, PaddlePaddle, and far too many to list — Additionally the software radio world is rich in open source tools surrounding the GNU Radio Ecosystem, which provides numerous tools for signal processing, modem simulation, channel simulation, etc. The sum of all of these can be a steep learning curve so we’ve attempted to curate some resources here to make getting up and running easier.
A docker image based on the nvidia-docker ubuntu base, including GNU Radio, Theano, TensorFlow, Keras and a handful of other tools to help provide foundations needed for work in this area. See Docker Image