search

amanchadha / iSeeBetter

iSeeBetter: Spatio-Temporal Video Super Resolution using Recurrent-Generative Back-Projection Networks | Python3 | PyTorch | GANs | CNNs | ResNets | RNNs | Published in Springer Journal of Computational Visual Media, September 2020, Tsinghua University Press
https://arxiv.org/abs/2006.11161
MIT License
359 stars 68 forks source link

Not Upscaling - getting the same resolution as input #21

Closed nigh8w0lf closed 3 years ago

nigh8w0lf commented 3 years ago

Using the --upscale_only switch, getting same resolution of images as input.

SolomGfxcolor commented 3 years ago

give me notebook for google colab Plz

spigelli commented 3 years ago

give me notebook for google colab Plz

A notebook to address the --upscale_only flag?

spigelli commented 3 years ago

@tthg119 @amanchadha Any news on getting this fixed? I'd like to use this upscaler in a project I'm working on

AromaticJ commented 3 years ago

@spigelli Hi, in original code of "iSeeBetterTest.py" the parameter "--upscale_only" doesn't transformeted to "dataset.py" and "data.py"。 You can follow the steps below to fix this error iSeeBetterTest.py revise:

  1. replace line38 with:parser.add_argument('-u', '--upscale_only', type=bool, default=True, help="Upscale mode - without downscaling.")
  2. repalce line56 with:test_set = get_test_set(args.data_dir, args.nFrames, args.upscale_factor, args.file_list, args.other_dataset, args.future_frame, args.upscale_only)

data.py revise:

  1. replace line24 with:def get_test_set(data_dir, nFrames, upscale_factor, file_list, other_dataset, future_frame, upscale_only):

  2. replace line25 with:return DatasetFromFolderTest(data_dir, nFrames, upscale_factor, file_list, other_dataset, future_frame, transform=transform(), upscale_only=upscale_only)

amanchadha commented 3 years ago

Thanks @AromaticJ for the fixes!

@spigelli , @nigh8w0lf : I've just pushed a commit with the fixes. Note that you will need to pass in -u for the upscale only mode to work (iSeeBetter is not set to 'upscale only' by default).