Ali Darwish, Ph.D.

You aren't using any samples for validation, where do you see that you are validation on 32 samples?

@fortunto2 It takes me to long to train on a Titan X! Approximately 1 second per image for a data set with 6 classes.

You may want to close this issue then, thanks!

Same thing here, ImportError: No module named vision.

I have the same issue. I have Nvidia 1080tis and 2080ti. The problem only occurs when I use the 2080ti for training but same code works perfectly fine on 1080ti....

I guess we need more reviewers. I can volunteer to review and test.

@CaryKnoop the problem is there isn't a way to compare before-after to see how the algorithm works. In fact there is no mention that TecoGan was used for the restoration.

@akirakitamura19 is that a normalization layer that you are talking about?

Hi @yera217 , have you figured that out?

@kshitijgoel007 a work around is to write a function that changes the state of the drone/car when a collision is detected. You can easly do that by using the info...