Fast Training of Convolutional Networks through FFTs · HackerPapers discussions

Fast Training of Convolutional Networks through FFTs — mentioned by hadeson

Mon, 28 Sep 2020 08:40:37 GMT

It could be used to accelerate Convolutional Neural Nets training [0] [0] https://arxiv.org/abs/1312.5851

Paper and discussion history. A mention is not an endorsement.

Fast Training of Convolutional Networks through FFTs — mentioned by adapteva

Mon, 25 May 2015 22:24:01 GMT

It will be an image processing demo. Here's a hint... http://arxiv.org/abs/1312.5851

Paper and discussion history. A mention is not an endorsement.

Fast Training of Convolutional Networks through FFTs — mentioned by im3w1l

Tue, 21 Apr 2015 01:18:47 GMT

You should use FFT's for the convolution, see http://arxiv.org/abs/1312.5851 Also, while the original Blas was in Fortran, you probably want OpenBlas, which is in C. Or maybe Intels MKL.

Paper and discussion history. A mention is not an endorsement.