I need an expert in Tensorflow and deep learning to implement the attached paper in Tensorflow. I have access to the data (RGB images, magnitude images, phase images, and the 5-fold text files with path to the images) and the network and solver prototexts in Caffe. You will need to convert the data to a format usable by Tensorflow (preferably TFrecords) and setup the network architecture in TF and train and test the network. I have done some preliminary work and I can share my code with you. It should be straightforward if you are an expert in Tensorflow and know how to deal with large size data (120*120*107). Basically you need to convert the code and data from Caffe to Tensorflow, and train the network and make sure you can replicate the results. This project has an optional part 2 that comes after successfully implementing part 1. I am setting up a maximum $250 budget for this. I am expecting you can finish this in less that 2 weeks. I am somehow flexible with the amount and time. Thanks!