Transform Quantization for CNN Compression

The dataset used in this paper is a collection of convolutional neural network (CNN) weights, which are compressed using transform quantization.

BibTex: