Network Compression: Skills

NETWORK COMPRESSION

VARIATIONAL CONVOLUTIONAL NEURAL NETWORK PRUNING

We propose a variational Bayesian scheme for pruning convolutional neural networks in channel level. In a nutshell, variational technique is introduced to estimate distribution of a newly proposed parameter, called channel saliency, based on this, redundant channels can be removed from model via a simple criterion.

Code

PERFORMANCE GUARANTEED NETWORK ACCELERATION VIA HIGH-ORDER RESIDUAL QUANTIZATION

In this paper, we propose a high-order binarization scheme, which achieves more accurate approximation while still possesses the advantage of binary operation. In particular, the proposed scheme recursively performs residual quantization and yields a series of binary input images with decreasing magnitude scales.

Code