Yuheng Qiu

I thought the first algorithm is implemented for cross entropy, which `discriminator_loss_fake + discriminator_loss_real`

It's this problem solved? I met a similar problem `4 errors detected in the compilation of "/tmp/tmpxft_000071e7_00000000-6_data_augmentation.cu.cpp1.ii". Makefile:62: recipe for target 'preprocessing' failed make: *** [preprocessing] Error 1`

@Iamanorange based on the previous post, I change the makefile to `F_INC = `python -c "import tensorflow; print(tensorflow.sysconfig.get_include())"` ifndef CUDA_HOME CUDA_HOME := /usr/local/cuda endif CC = gcc -O2 -pthread CXX...

Modify Makefile: TF_LIB = `python -c "import tensorflow; print(tensorflow.sysconfig.get_lib())"` CGPUFLAGS = -L$(CUDA_HOME)/lib -L$(CUDA_HOME)/lib64 -lcudart -L$(TF_LIB) -ltensorflow_framework I do the `make file` again. Traceback (most recent call last): File "/home/groupxghan/cv/.conda/envs/flow-tf/lib/python2.7/runpy.py", line...

Thanks a lot! I have solved this problem. I will be careful next time !!!!

@GRF-Sunomikp31 Thank you for pointing out, I really appreciate that. I will let you know when I update the paper and the code. The dataset has been uploaded. If you're...

Hi @sien110 I am still working on the code. I just released the dataset TartanAir Shibuya. If you're interested, you may take a look.

Same issue, with pytorch 0.4.1