advimman / deep-landscape Goto Github PK

Thanks for uploading comprehensive code. Unfortunately, I obtain only black images when using: PYTHONPATH=pwd:$PYTHONPATH runfiles/encode_and_animate_test_all256_with_style.sh <homography_dir> with omographies/manual_homographies_x2.5. I went through the code and found out that the function gen_images_cycle_shift produces images with values of -inf. Could you please tell me how to fix it? I followed all the steps given, install requirements, download model etc. ..

Hi can you please add a google colab for inference thanks

I read your paper and I think that your work is really interesting, and I'd really like to see your implementation.
When will you release them?

Is a GPU necessary to run the pre-trained models as well?

Hi thanks for releasing your code. Please could you add the version of PyTorch the requirements as I am getting conflicting dependency issues.
Thanks.

Hi, I checked out the paper and the testing code and found it is an interesting work.
But there is no training data to reproduce the same pretrained models.
I wonder if you have any plan to release the training data including static images and video lists(as described in the paper)?

Hi guys, congratulations for the wonderful work! I have seen you work via ECCV 2020. The idea is brilliant. I hope the source code will be released soon so that we can try it on multiple datasets. Are you planing to share it?