AliaksandrSiarohin

It is here https://yadi.sk/d/Xy9Pi-nF8jlPgw?w=1

See #15

Seems like a problem of new tensroflow version. Instead running on concatenated tensor (https://github.com/AliaksandrSiarohin/gan/blob/a4aa9a38792e89480ffbb15dc7e9beba8513e893/inception_score.py#L43) you can run them one by one in a for loop. ``` pred = [sess.run(softmax, {'ExpandDims:0':...

Hi. You mean for part swapping? No for part swapping you need segmentation. Fomm does not provide you with the segmentation.

Yes you can use it in supervised mode.

How much gpu memore you have? Try to run on cpu, by removing all calls to cuda.

See part_swap.ipynb for visualization.

First of all, test/****.png is actually videos. The frames are stacked together for a simpler i/o. 1. What do you mean resize? You resize some of your images to 64x64,...

1. Yes the keypoints is learned to be extracted at resolution 64x64, I doubt it generalize to higher resolutions. Need another model, or model trained on different resolutions. Model params...

What do you mean a segmentation based swapping? This work uses segmentation.