Visualizing results for the 1000 RAW test images?

yuanming-hu / exposure

Learning infinite-resolution image processing with GAN and RL from unpaired image datasets, using a differentiable photo editing model.

MIT License

767 stars 156 forks source link

Visualizing results for the 1000 RAW test images? #20

Closed thelittlekid closed 6 years ago

thelittlekid commented 6 years ago

Hi, @yuanming-hu. Thanks for sharing the codes. I'm wondering if there is an easy way to visualize the results for those 1000 RAW test images (part 3 of the MIT-Adobe FiveK Dataset). Thanks.

yuanming-hu commented 6 years ago

Thanks for the question. I will deal with it tonight.

yuanming-hu commented 6 years ago

Hi,

The results are here: https://github.com/yuanming-hu/exposure_models/releases/download/v0.0.1/test_outputs.zip

To generate them yourself, please download the package of test input images: https://github.com/yuanming-hu/exposure_models/releases/download/v0.0.1/test_set.zip and run

python3 evaluate.py example pretrained data/fivek_dataset/test_set/*.tif

and wait for a while (~60 minutes, seemingly bounded by some non-GPU computation which I haven't got a time to figure out yet -> #21 ).

thelittlekid commented 6 years ago

Thanks!