OctaveConv
OctaveConv copied to clipboard
terrychenism
A MXNet Implementation for Drop an Octave
First of all, feel the author's open source spirit, great idea! I know that low and high frequencies are divided according to a given proportion, but how are they divided...
On super-resolution tasks, the ordinary convolutional layer is replaced by octave convolution, but the loss and calculation time become larger, what is the matter?
dear author, how to deal with depthwise conv replaced by octave conv at last octave conv, the num_filter and num_group doesn't match, how to merge the high_f channel and low_f...
Thank you very much. I would welcome if you can share the 1d version of OctConv as well.
