context-encoder
context-encoder copied to clipboard
Implementation in pytorch
Any idea on how to implement this paper in pytorch would be of a great help.! And what do these line of the code do? How do I implement these lines in pytorch
real_ctx[{{},{1},{},{}}][mask_global] = 2 * 117.0/255.0 - 1.0 real_ctx[{{},{2},{},{}}][mask_global] = 2 * 104.0/255.0 - 1.0 real_ctx[{{},{3},{},{}}][mask_global] = 2 * 123.0/255.0 - 1.0 input_ctx:copy(real_ctx)
There is a GitHub repository, you may get some insights from that. But there are some differences with the original code (tt least, num_epochs and reconstruction loss differ)