Ding, Xiaohan, Ding, Guiguang, Guo, Yuchen, Han, Jungong and Yan, Chenggang (2019) Approximated oracle filter pruning for destructive CNN width optimization. In: ICML 2019 ; 36th International Conference on Machine Learning, California, 10-15 Jun 2019. Published in: Proceedings of the 36th International Conference on Machine Learning

Preview

PDF WRAP-approxiamated-oracle-filter-pruning-destructive-Han-2019.pdf - Accepted Version - Requires a PDF viewer. Download (1400Kb) | Preview

Request Changes to record.

Abstract

It is not easy to design and run Convolutional Neural Networks (CNNs) due to: 1) finding the optimal number of filters (i.e., the width) at each layer is tricky, given an architecture; and 2) the computational intensity of CNNs impedes the deployment on computationally limited devices. Oracle Pruning is designed to remove the unimportant filters from a well-trained CNN, which estimates the filters’ importance by ablating them in turn and evaluating the model, thus delivers high accuracy but suffers from intolerable time complexity, and requires a given resulting width but cannot automatically find it. To address these problems, we propose Approximated Oracle Filter Pruning (AOFP), which keeps searching for the least important filters in a binary search manner, makes pruning attempts by masking out filters randomly, accumulates the resulting errors, and finetunes the model via a multi-path framework. As AOFP enables simultaneous pruning on multiple layers, we can prune an existing very deep CNN with acceptable time cost, negligible accuracy drop, and no heuristic knowledge, or re-design a model which exerts higher accuracy and faster inference

Item Type:	Conference Item (Paper)
Alternative Title:
Subjects:	Q Science > QA Mathematics > QA76 Electronic computers. Computer science. Computer software
Divisions:	Faculty of Science > WMG (Formerly the Warwick Manufacturing Group)
Library of Congress Subject Headings (LCSH):	Neural networks (Computer science), Computer vision, Machine learning, Artificial intelligence -- Computer programs
Journal or Publication Title:	Proceedings of the 36th International Conference on Machine Learning
Official Date:	June 2019
Dates:	Date Event June 2019 Published
Date of first compliant deposit:	25 June 2019
Status:	Peer Reviewed
Publication Status:	Published
Access rights to Published version:	Open Access
RIOXX Funder/Project Grant:	Project/Grant ID RIOXX Funder Name Funder ID 2018YFC0807500 National Basic Research Program of China (973 Program) UNSPECIFIED 61571269 [NSFC] National Natural Science Foundation of China http://dx.doi.org/10.13039/501100001809 BX20180172 National Postdoctoral Program for Innovative Talents UNSPECIFIED 2018M640131 China Postdoctoral Science Foundation http://dx.doi.org/10.13039/501100002858
Conference Paper Type:	Paper
Title of Event:	ICML 2019 ; 36th International Conference on Machine Learning
Type of Event:	Conference
Location of Event:	California
Date(s) of Event:	10-15 Jun 2019
Open Access Version:	ArXiv

Request changes or add full text files to a record

Repository staff actions (login required)

View Item

Downloads