fork from https://github.com/junyanz/pytorch-CycleGAN-and-pix2pix
a.sh is train cmd
Video2XYZ 已经实现： [vid2dep demo show] [vid2vid demo show]
[video generate demo: children]
正在推进： vid2sped vid2angle vid2action vid2reward vid2nlp ref: createamind#3
更完善的期望是sensorfuse2representation, representation2XYZ, SF2Rpstt2XYZ,
对于progressGAN--pix2pixHD想法是 can train pred angle /2(左右) /4(大左小左 大右小右) /8 图像的越来越精细到功能训练的准确度越来越高。
!Curiosity-driven Exploration by Self-supervised Prediction