/freqlearning

Primary LanguagePython

Frequency learning

This library allows to work with images in frequency domain and after that return to the spatial domain.

Available conversions in the frequency spectrum:

  • LinearFourier2d
  • GeneralFourier2d

Usage

Example:

import torch
import torch.nn as nn

from freqlearning import LinearFourier2d
from freqlearning import GeneralFourier2d


input_image = torch.randn(1, 256, 256) # input shape (1, 256, 256)

model = MyModel()
#gafl_layer = LinearFourier2d((1, 256, 256))
gafl_layer = GeneralFourier2d((1, 256, 256))

gafl_model = nn.Sequential(
    gafl_layer,
    model
)

PyPI version:

$ pip install freqlearning

Latest version from source:

$ pip install git+https://github.com/YaroslavBespalov/freqlearning/freqlearning

Requirements

Installing requirements by pip:

pip3 install --upgrade -r ./requirements.txt

Contributors

  • Yaroslav Bespalov & Viktor Shipitsin.