AliaksandrSiarohin

Results 255 comments of AliaksandrSiarohin

@pidginred can you provide full stack trace, and your configs.

I don't know. I also used moving-gif dataset. You may coolect the dataset yourself.

Set num_repeats to 1.

Some dataset have short epochs only few steps, to make epochs longer this parameter is used.

Probably the headphones produce some problems. In general the is no much control over the output, just ```--adapt_scale``` so you may try without it.

You can not separate kpdetector from the rest. The most important aspect is the fact that network is able to reconstruct the driving image from the driving keypoints, so perceptual...

All of them affect keypoint training.

It is possible, because some people make an mobile app out of this. But I'm not an expert in runtime optimization, so could not tell. Everything that you suggest could...