Deep-Image-Analogy-PyTorch
Deep-Image-Analogy-PyTorch copied to clipboard
Double check with run-time
Hi, thanks for this great pytorch implementation. I run the code for the demo image pair on GPU and it takes about 40min. The result is good. Just double check, is my run-time close to yours?
Yes. I ran the code on VM of GCP which has a 4-core CPU and a K80, and it takes around 30 mins. The patchmatch process on low level features (large size) is quite slow because it does not run on GPU
@Ben-Louis Thanks for giving credit for my implementation. I would like to let you know that I have also released a GPU version of PatchMatch , written in https://github.com/harveyslash/PatchMatch
@harveyslash Do you have a GPU version of PatchMatch which works with your PyTorch version of Deep Image Analogy?
Yes. It is a drop in replacement for the cpu version
@harveyslash Thanks for your implementation for the GPU version. Remind that anyone must normalize your own images before using the PatchMatch algorithm.
However, using the GPU version PatchMatch, the result is a little different from the CPU version. I am still debugging on it.
updates?