This project extends the idea of the innovative architecture of Kolmogorov-Arnold Networks (KAN) to the Convolutional Layers, changing the classic linear transformation of the convolution to learnable non linear activations in each pixel.
MIT License
694
stars
68
forks
source link
[feat] Is it possible to run it with non-square kernels? #11
It is needed to make some small changes in kan_convolutional/convolution.py, changing kernel_side:int for a kernel shape. You can make a PR and we'll check it out
It is needed to make some small changes in kan_convolutional/convolution.py, changing kernel_side:int for a kernel shape. You can make a PR and we'll check it out