This Jupyter Notebook demonstrates how to accelerate 2D convolution using the Fast Fourier Transform (FFT). FFT-based convolution is particularly useful for large convolutional filters and input images.
This project is licensed under the MIT License. Feel free to explore and modify the code as a learning exercise.
This notebook was composed by René Chenard, a computer scientist and mathematician with a degree from Université Laval.
You can contact the author at: [email protected]