pith. sign in

arxiv: 1811.02639 · v1 · pith:HUXZYRTMnew · submitted 2018-10-29 · 💻 cs.CV

Demystifying Neural Network Filter Pruning

classification 💻 cs.CV
keywords filterfunctionalitypruningfiltersretrainingmagnitudemethodmethods
0
0 comments X
read the original abstract

Based on filter magnitude ranking (e.g. L1 norm), conventional filter pruning methods for Convolutional Neural Networks (CNNs) have been proved with great effectiveness in computation load reduction. Although effective, these methods are rarely analyzed in a perspective of filter functionality. In this work, we explore the filter pruning and the retraining through qualitative filter functionality interpretation. We find that the filter magnitude based method fails to eliminate the filters with repetitive functionality. And the retraining phase is actually used to reconstruct the remained filters for functionality compensation for the wrongly-pruned critical filters. With a proposed functionality-oriented pruning method, we further testify that, by precisely addressing the filter functionality redundancy, a CNN can be pruned without considerable accuracy drop, and the retraining phase is unnecessary.

This paper has not been read by Pith yet.

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.