Recognition Effects of Deep Convolutional Neural Network on Smudged Handwritten Digits

被引：0

作者：

Xu, Zhe ^{[1
,2
]}

Terada, Yusuke ^{[2
]}

Jia, Dongbao ^{[2
,3
]}

Cai, Zonghui ^{[2
]}

Gao, Shangce ^{[2
]}

机构：

[1] Changzhou Inst Technol, Sch Comp Informat & Engn, Changzhou 213032, Jiangsu, Peoples R China

[2] Huaihai Inst Technol, Sch Comp Engn, Lianyugang 222005, Peoples R China

[3] Univ Toyama, Fac Engn, Toyama 9308555, Japan

来源：

2018 5TH INTERNATIONAL CONFERENCE ON INFORMATION SCIENCE AND CONTROL ENGINEERING (ICISCE 2018) | 2018年

关键词：

Deep convolutional neural network; handwritten digits discrimination; MNIST dataset; noise; MODEL;

D O I：

10.1109/ICISCE.2018.00093

中图分类号：

TP [自动化技术、计算机技术];

学科分类号：

0812 ;

摘要：

Deep convolutional neural network (CNN) is known to be the first truly successful deep learning approach for image processing and understanding, e.g., the handwritten digits discrimination. However, in real applications such as handwritten zip code recognition, the collected images are commonly with smudged background. In this paper, we study the recognition effects of CNN on smudged digits and compared the results with three-layered perceptron. Experimental results based on MNIST dataset with smudged background (simulated by salt-and-pepper and gaussian noises) show that a drastic decline of recognition accuracy is observed for CNN, suggesting that the extracted features by convolutional operation and max pooling is very sensitive to the noise.

引用

页码：412 / 416

页数：5

共 50 条

[1] Fuzzy Logic Module of Convolutional Neural Network for Handwritten Digits Recognition
Popko, E. A.
Weinstein, I. A.
[J]. 5TH INTERNATIONAL CONFERENCE ON MATHEMATICAL MODELING IN PHYSICAL SCIENCES (IC-MSQUARE 2016), 2016, 738
[2] Recognition of Bengali Handwritten Digits Using Convolutional Neural Network Architectures
Hasan, Md Mahmudul
Ul Islam, Md Rafid
Mahmood, Md Tareq
[J]. 2018 INTERNATIONAL CONFERENCE ON BANGLA SPEECH AND LANGUAGE PROCESSING (ICBSLP), 2018,
[3] DeepBanglaNet: A Deep Convolutional Neural Network to Recognize Bengali Handwritten Digits
Mahmud, Tanvir
Hossain, Abdul Rakib
Fattah, Shaikh Anowarul
[J]. 2020 IEEE REGION 10 SYMPOSIUM (TENSYMP) - TECHNOLOGY FOR IMPACTFUL SUSTAINABLE DEVELOPMENT, 2020, : 742 - 745
[4] Wavelet Convolutional Neural Networks for Handwritten Digits Recognition
Ben Chaabane, Chiraz
Mellouli, Dorra
Hamdani, Tarek M.
Alimi, Adel M.
Abraham, Ajith
[J]. HYBRID INTELLIGENT SYSTEMS, HIS 2017, 2018, 734 : 305 - 310
[5] Handwritten Hindi Digits Recognition Using Convolutional Neural Network with RMSprop Optimization
Reddy, R. Vijaya Kumar
Rao, B. Srinivasa
Raju, Prudvi
[J]. PROCEEDINGS OF THE 2018 SECOND INTERNATIONAL CONFERENCE ON INTELLIGENT COMPUTING AND CONTROL SYSTEMS (ICICCS), 2018, : 45 - 51
[6] Handwritten Digits Recognition With Subgroup Neural Network
Wang Minghui
Pan Xinan
Zhong Yixin (Department of Information Engineering
[J]. The Journal of China Universities of Posts and Telecommunications, 1994, (01) : 52 - 56
[7] Handwritten Digits Recognition with Artificial Neural Network
Islam, Kh Tohidul
Mujtaba, Ghulam
Raj, Ram Gopal
Nweke, Henry Friday
[J]. 2017 INTERNATIONAL CONFERENCE ON ENGINEERING TECHNOLOGY AND TECHNOPRENEURSHIP (ICE2T), 2017,
[8] A Deep Convolutional Neural Network for Bangla Handwritten Numeral Recognition
Islam, Kazi Mejbaul
Noor, Rouhan
Saha, Chaity
Rahimi, Md Jakaria
[J]. 2018 4TH IEEE INTERNATIONAL WIE CONFERENCE ON ELECTRICAL AND COMPUTER ENGINEERING (IEEE WIECON-ECE 2018), 2018, : 45 - 50
[9] Rejection strategy for convolutional neural network by adaptive topology applied to handwritten digits recognition
Cecotti, H
Belaïd, A
[J]. EIGHTH INTERNATIONAL CONFERENCE ON DOCUMENT ANALYSIS AND RECOGNITION, VOLS 1 AND 2, PROCEEDINGS, 2005, : 765 - 769
[10] Pashto isolated digits recognition using deep convolutional neural network
Zada, Bakht
Ullah, Rahim
[J]. HELIYON, 2020, 6 (02)

← 1 2 3 4 5 →