Sorry, you need to enable JavaScript to visit this website.

facebooktwittermailshare

ARCHITECTURE-AWARE NETWORK PRUNING FOR VISION QUALITY APPLICATIONS

Abstract: 

Convolutional neural network (CNN) delivers impressive achievements in computer vision and machine learning field. However, CNN incurs high computational complexity, especially for vision quality applications because of large image resolution. In this paper, we propose an iterative architecture-aware pruning algorithm with adaptive magnitude threshold while cooperating with quality-metric measurement simultaneously. We show the performance improvement applied on vision quality applications and provide comprehensive analysis with flexible pruning configuration. With the proposed method, the Multiply-Accumulate (MAC) of state-of-the-art low-light imaging (SID) and super-resolution (EDSR) are reduced by 58% and 37% without quality drop, respectively. The memory bandwidth (BW) requirements of convolutional layer can be also reduced by 20% to 40%.

up
0 users have voted:

Paper Details

Authors:
Wei-Ting Wang, Han-Lin Li, Wei-Shiang Lin, Cheng-Ming Chiang, Yi-Min Tsai
Submitted On:
12 September 2019 - 1:21am
Short Link:
Type:
Poster
Event:
Presenter's Name:
Wei-Ting Wang
Document Year:
2019
Cite

Document Files

ICIP_2019_poster_1686_ARCHITECTURE-AWARE_NETWORK_PRUNING_FOR_VISION_QUALITY_APPLICATIONS.pdf

(7)

Subscribe

[1] Wei-Ting Wang, Han-Lin Li, Wei-Shiang Lin, Cheng-Ming Chiang, Yi-Min Tsai, "ARCHITECTURE-AWARE NETWORK PRUNING FOR VISION QUALITY APPLICATIONS", IEEE SigPort, 2019. [Online]. Available: http://sigport.org/4599. Accessed: Sep. 15, 2019.
@article{4599-19,
url = {http://sigport.org/4599},
author = {Wei-Ting Wang; Han-Lin Li; Wei-Shiang Lin; Cheng-Ming Chiang; Yi-Min Tsai },
publisher = {IEEE SigPort},
title = {ARCHITECTURE-AWARE NETWORK PRUNING FOR VISION QUALITY APPLICATIONS},
year = {2019} }
TY - EJOUR
T1 - ARCHITECTURE-AWARE NETWORK PRUNING FOR VISION QUALITY APPLICATIONS
AU - Wei-Ting Wang; Han-Lin Li; Wei-Shiang Lin; Cheng-Ming Chiang; Yi-Min Tsai
PY - 2019
PB - IEEE SigPort
UR - http://sigport.org/4599
ER -
Wei-Ting Wang, Han-Lin Li, Wei-Shiang Lin, Cheng-Ming Chiang, Yi-Min Tsai. (2019). ARCHITECTURE-AWARE NETWORK PRUNING FOR VISION QUALITY APPLICATIONS. IEEE SigPort. http://sigport.org/4599
Wei-Ting Wang, Han-Lin Li, Wei-Shiang Lin, Cheng-Ming Chiang, Yi-Min Tsai, 2019. ARCHITECTURE-AWARE NETWORK PRUNING FOR VISION QUALITY APPLICATIONS. Available at: http://sigport.org/4599.
Wei-Ting Wang, Han-Lin Li, Wei-Shiang Lin, Cheng-Ming Chiang, Yi-Min Tsai. (2019). "ARCHITECTURE-AWARE NETWORK PRUNING FOR VISION QUALITY APPLICATIONS." Web.
1. Wei-Ting Wang, Han-Lin Li, Wei-Shiang Lin, Cheng-Ming Chiang, Yi-Min Tsai. ARCHITECTURE-AWARE NETWORK PRUNING FOR VISION QUALITY APPLICATIONS [Internet]. IEEE SigPort; 2019. Available from : http://sigport.org/4599