Iamanorange

@mengyaaa Should run Flownet2 on GPU only.

> but I got some problems when I install the python-tk in the Docker container, I try to use " apt-get install python-tk " TkInter (python-tk) is used to draw...

DO NOT COMMENT -DGOOGLE_CUDA=1, or you will lose GPU support. If your error mentioned: > error: calling a constexpr \_\_host__ function("xxxx") from a \_\_device__ function("CudaAtomicAdd") is not allowed. The experimental...

You error seems different from irwinherrmann's (you have only 1 error for target preprocessing but he has 48). Can you paste your full error message? The Makefile should be OK....

Add `-DNDEBUG` to CFLAGS. For more information: https://github.com/tensorflow/tensorflow/issues/22766

Congratulation. The only error is "constexpr function return is non-constant" and can be ignored by `-DNDEBUG`. The rest are just warnings, and can be ignored by everyone.

@leeskyed Use GPU.

https://github.com/sampepose/flownet2-tf/issues/28#issuecomment-406941839 Edit: You are using Python 2.6, seriously?

Inference 3 times, and merge them as you want.

https://github.com/sampepose/flownet2-tf/issues/45#issuecomment-411713153