Research on Traffic Sign Recognition Method Based on Transfer Learning
Article No.: 218, Pages 1 - 5
Abstract
In order to solve the problems in existing traffic sign recognition algorithms, this paper proposes a traffic sign recognition algorithm based on migration learning. This algorithm proposes to use the NSLC-AE architecture model to reconstruct clean data from dirty data (source and target data) To learn low-dimensional features that are robust to known natural noise (source noise and target noise). That is, in the low-dimensional space, use local constraints and discriminant information to find semantically similar samples for each sample, and construct clean data for model training in the high-dimensional space. Because in the low-dimensional space, high-dimensional natural noise can usually be decoupled into one or several hidden variables, while the interference characteristics in the low-dimensional space can simulate the influence of multiple natural noises. Therefore, this paper further proposes the N-GAN architecture based on the adversarial generative model, and trains the denoiser and mapper in the low-dimensional space in reverse to learn features that are robust to various natural noises. The experimental results show The improved model proposed in this paper can greatly reduce the image blur and the probability of recognition errors in the process of traffic sign recognition.
References
[1]
Chen Xiuxin, Ye Yang, Yu Zhong, Zhang Xue. Identification of Traffic Signs in Haze Weather Based on Deep Learning [J].]; and Journal of Chongqing Jiaotong University (Natural Science Edition), 2020,39(12):1-5+12.
[2]
Shen Yuan. Algorithm Research on Identification and Detection of Traffic Signs Zhejiang Institute of Science and Technology,2020.
[3]
CHEN M, XU Z, WEINBERGER K Q, Marginalized denoising autoencoders fordomain adaptation[C]//Proceedings of the 29th International Coference on Interna-tional Conference on Machine Learning. Omnipress. [S.l.]:[s.n.], 2012:1627–1634.
[4]
WANG C, MAHADEVAN S. Heterogeneous domain adaptation using manifold align-ment[C]//Twenty-Second International Joint Conference on Artificial Intelligence.Vol. 22. 1. [S.l.]: [s.n.], 2011:1541.
[5]
GOPALAN R, LI R, CHELLAPPA R. Domain adaptation for object recognition: Anunsupervised approach[C]//Computer Vision, 2011 IEEE International Conference on. IEEE. [S.l.]: [s.n.], 2011:999–1006.
[6]
LONG M, WANG J, DING G, Transfer feature learning with joint distribution adaptation[C]//Proceedings of the IEEE international conference on computer vision.[S.l.]: [s.n.], 2013:2200–2207.
[7]
DINGZ,SHAOM,FUY.Deep low-rank coding for transfer learning[C]//Twenty-Fourth International Joint Conference on Artificial Intelligence. [S.l.]: [s.n.], 2015.
[8]
Wang Kejun, Zhao Yandong, Xing Xianglei. Advances in the Application of Deep Learning in the Field of Unmanned Vehicles [J].]; and Intelligent Journal of Systems, 2018, 13(1):55–69.
[9]
Zhang. A Review of the Research on Traffic Signs Recognition Based on Deep Learning [J].1 Electronic world,2021(03):65-66.
[10]
GLOROT X, BORDES A, BENGIO Y. Domain adaptation for large-scale sentiment classification: A deep learning approach[G]//Proceedings of the 28th international conference on machine learning. [S.l.]: [s.n.], 2011:513–520.
[11]
Sea Song Hao. Fast Detection and Identification of Traffic Signs Based on Convolutional Neural Network [D]. Harbin Institute of Technology,2020.
[12]
DING Z, SHAO M, FU Y. Deep Robust Encoder Through Locality Preserving Low-Rank Dictionary[G]//European Conference on Computer Vision. [S.l.]: [s.n.], 2016:567–582.
[13]
BARÓ X, ESCALERA S, VITRIÀ J, Traffic sign recognition using evolutionary adaboost detection and forest-ECOC classification[J]. IEEE Transactions on Intelligent Transportation Systems, 2009, 10(1):113–126
Recommendations
Sparse-Representation-Based Graph Embedding for Traffic Sign Recognition
Researchers have proposed various machine learning algorithms for traffic sign recognition, which is a supervised multicategory classification problem with unbalanced class frequencies and various appearances. We present a novel graph embedding ...
Research on Traffic sign recognition based on CNN Deep Learning Network
AbstractWith the continuous development of social economy, the road traffic system as one of the basis of national economy is also increasingly developed, which is followed by the traffic safety problems are increasingly prominent. Among them, driving ...
Face recognition using discriminant locality preserving projections based on maximum margin criterion
In this paper, we propose a new discriminant locality preserving projections based on maximum margin criterion (DLPP/MMC). DLPP/MMC seeks to maximize the difference, rather than the ratio, between the locality preserving between-class scatter and ...
Comments
Information & Contributors
Information
Published In
May 2021
2053 pages
ISBN:9781450390200
DOI:10.1145/3469213
Copyright © 2021 ACM.
Permission to make digital or hard copies of all or part of this work for personal or classroom use is granted without fee provided that copies are not made or distributed for profit or commercial advantage and that copies bear this notice and the full citation on the first page. Copyrights for components of this work owned by others than ACM must be honored. Abstracting with credit is permitted. To copy otherwise, or republish, to post on servers or to redistribute to lists, requires prior specific permission and/or a fee. Request permissions from [email protected]
Publisher
Association for Computing Machinery
New York, NY, United States
Publication History
Published: 18 August 2021
Check for updates
Qualifiers
- Research-article
- Research
- Refereed limited
Conference
ICAIIS 2021
ICAIIS 2021: 2021 2nd International Conference on Artificial Intelligence and Information Systems
May 28 - 30, 2021
Chongqing, China
Contributors
Other Metrics
Bibliometrics & Citations
Bibliometrics
Article Metrics
- 0Total Citations
- 43Total Downloads
- Downloads (Last 12 months)10
- Downloads (Last 6 weeks)3
Reflects downloads up to 09 Nov 2024
Other Metrics
Citations
View Options
Get Access
Login options
Check if you have access through your login credentials or your institution to get full access on this article.
Sign inFull Access
View options
View or Download as a PDF file.
PDFeReader
View online with eReader.
eReaderHTML Format
View this article in HTML Format.
HTML Format