메뉴 건너뛰기
Library Notice
Institutional Access
If you certify, you can access the articles for free.
Check out your institutions.
ex)Hankuk University, Nuri Motors
Log in Register Help KOR
Subject
Humanities
General Humanities History Philosophy Religious Studies Linguistics and Literature Other Humanities
Social science
General Social Science Business Management Tourism Education Military Science Library and Information Science Law Sociology Journalism and Media Studies Psychological Science Political Science and Diplomacy Geography Public Administration Economics International Trade Anthropology Other Social Science
Natural science
General Natural Science Physics Biology Life Sciences Mathematics/Statistics Astronomy Earth Science Chemistry Other Natural Science
Engineering
General Engineering Architectural Engineering Mechanical Engineering Industrial Engineering Material/Energy Engineering Electrical and Electronic Engineering Marine and Ocean Engineering Chemical Engineering Computer Science Other Engineering
Medical science
General Medical Science Nursing Science Veterinary Pharmacy Traditional Korean Medicine Basic Medicine Clinical Medicine Other Medicine
Agriculture and Fishery
General Agriculture and Fishery Agriculture Fishery Science Food Science Forestry Animal Husbandry Marine Transportation Studies Other Agriculture and Fishery
Art and Kinesiology
General Arts Design Fine Arts Cosmetology Photography Music Fashion Kinesiology Performing Media Arts Other Arts and Kinesiology
Interdisciplinary Studies
Interdisciplinary Research Science and Technology Studies Brain and Cognitive Science Technology Policy Other Interdisciplinary Studies

KOR

Performance Analysis of Hint-KD Training Approach for the Teacher-Student Framework Using Deep Residual Networks
Recommendations
Search
Questions

딥 residual network를 이용한 선생-학생 프레임워크에서 힌트-KD 학습 성능 분석

논문 기본 정보

Type
Academic journal
Author
Ji-Hoon Bae (한국전자통신연구원) Junho Yim (한국과학기술원) Jaehak Yu (한국전자통신연구원) Kwihoon Kim (한국전자통신연구원) Junmo Kim (한국과학기술원)
Journal
The Institute of Electronics and Information Engineers Journal of the Institute of Electronics and Information Engineers Vol.54 No.5 (Wn.474) KCI Excellent Accredited Journal
Published
2017.5
Pages
35 - 41 (7page)

Usage

cover
📌
Topic
📖
Background
🔬
Method
🏆
Result
Performance Analysis of Hint-KD Training Approach for the Teacher-Student Framework Using Deep Residual Networks
Ask AI
Recommendations
Search
Questions

Abstract· Keywords

Report Errors
In this paper, we analyze the performance of the recently introduced Hint-knowledge distillation (KD) training approach based on the teacher-student framework for knowledge distillation and knowledge transfer. As a deep neural network (DNN) considered in this paper, the deep residual network (ResNet), which is currently regarded as the latest DNN, is used for the teacher-student framework. Therefore, when implementing the Hint-KD training, we investigate the impact on the weight of KD information based on the soften factor in terms of classification accuracy using the widely used open deep learning frameworks, Caffe. As a results, it can be seen that the recognition accuracy of the student model is improved when the fixed value of the KD information is maintained rather than the gradual decrease of the KD information during training.
#Knowledge distillation #Hint training #Deep residual networks #Caffe

Contents

요약
Abstract
Ⅰ. 서론
Ⅱ. 본론
Ⅲ. 실험 결과
Ⅳ. 결론
REFERENCES

References (12)

Add References
Y.-T. Park, “A comparative study of image recognition by neural network classifier and linear tree classifier”, Journal of The Institute of Electronics and Information Engineers-B, vol. 31, no. 5, pp. 141-148, 1994. S. Hong, W. Im, J. Park, and H.-S. Yang, “Deep CNN-based person identification using facial and clothing features”, in Proc. of Summer Conference on Institute of Electronics and Information Engineers (IEIE), pp. 2204-2207, June, 2016. Y. Shin, J.-H. Park, S. Shin, G. Lim, S. Song, C. Lee, and J.-M. Chung, “Improvement of image classification in augmented reality based on deep learning”, in Proc. of Summer Conference on Institute of Electronics and Information Engineers (IEIE), pp. 1771-1773, June, 2016. G. Hinton, O. Vinyals, and J. Dean, “Distilling the knowledge in a neural netwok”, arXiv prreprint arXiv:1503.02531, pp. 1-19, 2015. A. Romero, N. Ballas, S.E. Kahou, A. Chassang, C. Gatta, and Y. Bengio, “Fitnets: Hints for thin deep nets”, in Proc. of 5th International Conference on Learning Representations (ICLR), pp. 1-13, San Diego, May 7-9, 2015.

Recommendations

It is an article recommended by DBpia according to the article similarity. Check out the related articles!

Related Authors

Frequently Viewed Together

Face recognition Based on Super-resolution Method Using Sparse Representation and Deep Learning
Ohseol Kwon Journal of Korea Multimedia Society 2018 .02
Domestic Technical Standards and Performance Test of Photovoltaic PCS for Renewable Energy
Yong-Ho Yoon The transactions of The Korean Institute of Electrical Engineers 2018 .03
Deep learning based video streaming application using DASH
I-Seul Kim ,  Jun-Woo Mun ,  Kyeong-Jin Pa and others 5 한국HCI학회 학술대회 2018 .01
Fashion Trend Analysis Using Word Embedding
Jane E. Lee ,  Hwon Ihm ,  Soon-Bum Lim 한국HCI학회 학술대회 2018 .01
Learning music data with Recurrent neural network model
Esthe ,  Chung ,  Yongduek Seo 한국HCI학회 학술대회 2018 .01

Recently viewed articles

A Grounded Theoretical Study on the Process of Peer Relations Formation in Multi-cultural Family Students through Elementary Sports Club Activity
Baek， Jung-MinKorea Sport Society2020
 The Globalization of Taekwondo Recognized by Overseas Taekwondo Leaders
Jeon, Nan-Hee, KIM OK HYUN2019

Comments(0)

0

Write first comments.

UCI(KEPA) : I410-ECN-0101-2018-569-000888438