BPNet: Branch-pruned conditional neural network for systematic time-accuracy tradeoff

Kyungchul Park; Chanyoung Oh; Youngmin Yi

doi:10.1109/DAC18072.2020.9218545

BPNet: Branch-pruned conditional neural network for systematic time-accuracy tradeoff

Kyungchul Park
, Chanyoung Oh
, Youngmin Yi

School of Electrical and Computer Engineering

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

9 Scopus citations

Abstract

Recently, there have been attempts to execute the neural network conditionally with auxiliary classifiers allowing early termination depending on the difficulty of the input, which can reduce the execution time or energy consumption without any or with negligible accuracy decrease. However, previous studies do not consider how many or where the auxiliary classifiers, or branches, should be added in a systematic fashion. In this paper, we propose Branch-pruned Conditional Neural Network (BPNet) and its methodology in which the time-accuracy tradeoff for the conditional neural network can be found systematically. We applied BPNet to SqueezeNet, ResNet-20, and VGG-16 with CIFAR-10 and 100. BPNet achieves up to 3.15× of speedup without any accuracy drop compared to the base networks.

Original language	English
Title of host publication	2020 57th ACM/IEEE Design Automation Conference, DAC 2020
Publisher	Institute of Electrical and Electronics Engineers Inc.
ISBN (Electronic)	9781450367257
DOIs	https://doi.org/10.1109/DAC18072.2020.9218545
State	Published - Jul 2020
Event	57th ACM/IEEE Design Automation Conference, DAC 2020 - Virtual, San Francisco, United States Duration: 20 Jul 2020 → 24 Jul 2020

Publication series

Name	Proceedings - Design Automation Conference
Volume	2020-July
ISSN (Print)	0738-100X

Conference

Conference	57th ACM/IEEE Design Automation Conference, DAC 2020
Country/Territory	United States
City	Virtual, San Francisco
Period	20/07/20 → 24/07/20

UN SDGs

This output contributes to the following UN Sustainable Development Goals (SDGs)

SDG 7 Affordable and Clean Energy

Keywords

And design space exploration
Conditional neural networks
Deep neural networks (DNNs)
Embedded systems

Access to Document

10.1109/DAC18072.2020.9218545

Cite this

@inproceedings{a45b6fde9cd7456d9027d2861a42126c,

title = "BPNet: Branch-pruned conditional neural network for systematic time-accuracy tradeoff",

abstract = "Recently, there have been attempts to execute the neural network conditionally with auxiliary classifiers allowing early termination depending on the difficulty of the input, which can reduce the execution time or energy consumption without any or with negligible accuracy decrease. However, previous studies do not consider how many or where the auxiliary classifiers, or branches, should be added in a systematic fashion. In this paper, we propose Branch-pruned Conditional Neural Network (BPNet) and its methodology in which the time-accuracy tradeoff for the conditional neural network can be found systematically. We applied BPNet to SqueezeNet, ResNet-20, and VGG-16 with CIFAR-10 and 100. BPNet achieves up to 3.15× of speedup without any accuracy drop compared to the base networks.",

keywords = "And design space exploration, Conditional neural networks, Deep neural networks (DNNs), Embedded systems",

author = "Kyungchul Park and Chanyoung Oh and Youngmin Yi",

note = "Publisher Copyright: {\textcopyright} 2020 IEEE.; 57th ACM/IEEE Design Automation Conference, DAC 2020 ; Conference date: 20-07-2020 Through 24-07-2020",

year = "2020",

month = jul,

doi = "10.1109/DAC18072.2020.9218545",

language = "English",

series = "Proceedings - Design Automation Conference",

publisher = "Institute of Electrical and Electronics Engineers Inc.",

booktitle = "2020 57th ACM/IEEE Design Automation Conference, DAC 2020",

address = "United States",

}

Park, K, Oh, C & Yi, Y 2020, BPNet: Branch-pruned conditional neural network for systematic time-accuracy tradeoff. in 2020 57th ACM/IEEE Design Automation Conference, DAC 2020., 9218545, Proceedings - Design Automation Conference, vol. 2020-July, Institute of Electrical and Electronics Engineers Inc., 57th ACM/IEEE Design Automation Conference, DAC 2020, Virtual, San Francisco, United States, 20/07/20. https://doi.org/10.1109/DAC18072.2020.9218545

BPNet: Branch-pruned conditional neural network for systematic time-accuracy tradeoff. / Park, Kyungchul; Oh, Chanyoung; Yi, Youngmin.
2020 57th ACM/IEEE Design Automation Conference, DAC 2020. Institute of Electrical and Electronics Engineers Inc., 2020. 9218545 (Proceedings - Design Automation Conference; Vol. 2020-July).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

TY - GEN

T1 - BPNet

T2 - 57th ACM/IEEE Design Automation Conference, DAC 2020

AU - Park, Kyungchul

AU - Oh, Chanyoung

AU - Yi, Youngmin

PY - 2020/7

Y1 - 2020/7

N2 - Recently, there have been attempts to execute the neural network conditionally with auxiliary classifiers allowing early termination depending on the difficulty of the input, which can reduce the execution time or energy consumption without any or with negligible accuracy decrease. However, previous studies do not consider how many or where the auxiliary classifiers, or branches, should be added in a systematic fashion. In this paper, we propose Branch-pruned Conditional Neural Network (BPNet) and its methodology in which the time-accuracy tradeoff for the conditional neural network can be found systematically. We applied BPNet to SqueezeNet, ResNet-20, and VGG-16 with CIFAR-10 and 100. BPNet achieves up to 3.15× of speedup without any accuracy drop compared to the base networks.

AB - Recently, there have been attempts to execute the neural network conditionally with auxiliary classifiers allowing early termination depending on the difficulty of the input, which can reduce the execution time or energy consumption without any or with negligible accuracy decrease. However, previous studies do not consider how many or where the auxiliary classifiers, or branches, should be added in a systematic fashion. In this paper, we propose Branch-pruned Conditional Neural Network (BPNet) and its methodology in which the time-accuracy tradeoff for the conditional neural network can be found systematically. We applied BPNet to SqueezeNet, ResNet-20, and VGG-16 with CIFAR-10 and 100. BPNet achieves up to 3.15× of speedup without any accuracy drop compared to the base networks.

KW - And design space exploration

KW - Conditional neural networks

KW - Deep neural networks (DNNs)

KW - Embedded systems

UR - https://www.scopus.com/pages/publications/85093927085

U2 - 10.1109/DAC18072.2020.9218545

DO - 10.1109/DAC18072.2020.9218545

M3 - Conference contribution

AN - SCOPUS:85093927085

T3 - Proceedings - Design Automation Conference

BT - 2020 57th ACM/IEEE Design Automation Conference, DAC 2020

PB - Institute of Electrical and Electronics Engineers Inc.

Y2 - 20 July 2020 through 24 July 2020

ER -

BPNet: Branch-pruned conditional neural network for systematic time-accuracy tradeoff

Abstract

Publication series

Conference

UN SDGs

Keywords

Access to Document

Other files and links

Fingerprint

Cite this