LAFD-Net: Learning with Noisy Pseudo Labels for Semi-Supervised Bearing Fault Diagnosis
Fault Diagnosis for the rolling bearing is an important field that has received increasing attention in recent years. The main challenge for this task is the lack of labeled data. Existing works circumvent this problem with pseudo labels generated from labeled data. However, these pseudo labels are...
Gespeichert in:
Veröffentlicht in: | IEEE sensors journal 2023-01, p.1-1 |
---|---|
Hauptverfasser: | , , , , , , , |
Format: | Artikel |
Sprache: | eng |
Schlagworte: | |
Online-Zugang: | Volltext bestellen |
Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
container_end_page | 1 |
---|---|
container_issue | |
container_start_page | 1 |
container_title | IEEE sensors journal |
container_volume | |
creator | Jian, Yifan Chen, Zhi Lei, Yinjie He, Zhengxi Zhao, Yang He, Liang Luo, Wei Chen, Xuekun |
description | Fault Diagnosis for the rolling bearing is an important field that has received increasing attention in recent years. The main challenge for this task is the lack of labeled data. Existing works circumvent this problem with pseudo labels generated from labeled data. However, these pseudo labels are noisy even with consistency check or confidence-based filtering due to the minimal amount of training data. To solve this problem, a novel Label-level Anti-noise Fault Diagnosis Network (LAFD-Net) is proposed in this paper. Specifically, we propose an Online Asymptotic Label Updating (OALU) strategy which contains two updating stages: self-correction stage and cross-correction stage. The proposed OALU can stably and reliably generate new corrected pseudo labels, gradually replacing the old noisy ones. The LAFD-Net adopts a Student-Teacher architecture. For such a Student-Teacher model, we propose a Consistency Enhancement (CE) loss to strengthen the feature consistency between student and teacher networks, aiming to achieve more efficient use of plentiful unlabeled data via feature regularization. Finally, a series of experiments were conducted using the University of Cincinnati Intelligent Maintenance System (IMS) Center dataset and the Case Western Reserve University (CWRU) bearing dataset. The experimental results demonstrate that the proposed semi-supervised learning schemes outperformed existing state-of-the-art methods with the same percentage of labeled data samples. |
doi_str_mv | 10.1109/JSEN.2023.3233957 |
format | Article |
fullrecord | <record><control><sourceid>ieee_RIE</sourceid><recordid>TN_cdi_ieee_primary_10014694</recordid><sourceformat>XML</sourceformat><sourcesystem>PC</sourcesystem><ieee_id>10014694</ieee_id><sourcerecordid>10014694</sourcerecordid><originalsourceid>FETCH-ieee_primary_100146943</originalsourceid><addsrcrecordid>eNqFjsuKwkAQAOegYHx8gOChfyDZmUxiyN58hUVCECLoTUbSai8xkelE8e9V2Pue6lAUlBBjJT2lZPy1zleZ50tfe9rXOg6jjnBUqKUb6GjfE33mXylVHIWRI3bpLFm6GTbfkKKxFVVneFBzgawmfsKGsS1qSM0RS4ZTbSHHK7l5e0N7J8YC5u_qEyWmLRtYkjlXNRMPRfdkSsbRHwdikqy2ix-XEPFws3Q19nlQ749gGgf6H_0C8FI_tg</addsrcrecordid><sourcetype>Publisher</sourcetype><iscdi>true</iscdi><recordtype>article</recordtype></control><display><type>article</type><title>LAFD-Net: Learning with Noisy Pseudo Labels for Semi-Supervised Bearing Fault Diagnosis</title><source>IEEE Electronic Library (IEL)</source><creator>Jian, Yifan ; Chen, Zhi ; Lei, Yinjie ; He, Zhengxi ; Zhao, Yang ; He, Liang ; Luo, Wei ; Chen, Xuekun</creator><creatorcontrib>Jian, Yifan ; Chen, Zhi ; Lei, Yinjie ; He, Zhengxi ; Zhao, Yang ; He, Liang ; Luo, Wei ; Chen, Xuekun</creatorcontrib><description>Fault Diagnosis for the rolling bearing is an important field that has received increasing attention in recent years. The main challenge for this task is the lack of labeled data. Existing works circumvent this problem with pseudo labels generated from labeled data. However, these pseudo labels are noisy even with consistency check or confidence-based filtering due to the minimal amount of training data. To solve this problem, a novel Label-level Anti-noise Fault Diagnosis Network (LAFD-Net) is proposed in this paper. Specifically, we propose an Online Asymptotic Label Updating (OALU) strategy which contains two updating stages: self-correction stage and cross-correction stage. The proposed OALU can stably and reliably generate new corrected pseudo labels, gradually replacing the old noisy ones. The LAFD-Net adopts a Student-Teacher architecture. For such a Student-Teacher model, we propose a Consistency Enhancement (CE) loss to strengthen the feature consistency between student and teacher networks, aiming to achieve more efficient use of plentiful unlabeled data via feature regularization. Finally, a series of experiments were conducted using the University of Cincinnati Intelligent Maintenance System (IMS) Center dataset and the Case Western Reserve University (CWRU) bearing dataset. The experimental results demonstrate that the proposed semi-supervised learning schemes outperformed existing state-of-the-art methods with the same percentage of labeled data samples.</description><identifier>ISSN: 1530-437X</identifier><identifier>DOI: 10.1109/JSEN.2023.3233957</identifier><identifier>CODEN: ISJEAZ</identifier><language>eng</language><publisher>IEEE</publisher><subject>Bearing fault ; Data models ; Deep learning ; Fault diagnosis ; Feature extraction ; learning with noisy data ; Noise measurement ; semi-supervised Learning ; Sensors ; student-teacher model ; Training</subject><ispartof>IEEE sensors journal, 2023-01, p.1-1</ispartof><lds50>peer_reviewed</lds50><woscitedreferencessubscribed>false</woscitedreferencessubscribed><orcidid>0000-0002-5820-9080 ; 0000-0002-2683-384X ; 0000-0002-2683-384X ; 0000-0002-8211-3046 ; 0000-0001-6856-3342 ; 0000-0002-2683-384X ; 0000-0002-2683-384X</orcidid></display><links><openurl>$$Topenurl_article</openurl><openurlfulltext>$$Topenurlfull_article</openurlfulltext><thumbnail>$$Tsyndetics_thumb_exl</thumbnail><linktohtml>$$Uhttps://ieeexplore.ieee.org/document/10014694$$EHTML$$P50$$Gieee$$H</linktohtml><link.rule.ids>314,780,784,796,27924,27925,54758</link.rule.ids><linktorsrc>$$Uhttps://ieeexplore.ieee.org/document/10014694$$EView_record_in_IEEE$$FView_record_in_$$GIEEE</linktorsrc></links><search><creatorcontrib>Jian, Yifan</creatorcontrib><creatorcontrib>Chen, Zhi</creatorcontrib><creatorcontrib>Lei, Yinjie</creatorcontrib><creatorcontrib>He, Zhengxi</creatorcontrib><creatorcontrib>Zhao, Yang</creatorcontrib><creatorcontrib>He, Liang</creatorcontrib><creatorcontrib>Luo, Wei</creatorcontrib><creatorcontrib>Chen, Xuekun</creatorcontrib><title>LAFD-Net: Learning with Noisy Pseudo Labels for Semi-Supervised Bearing Fault Diagnosis</title><title>IEEE sensors journal</title><addtitle>JSEN</addtitle><description>Fault Diagnosis for the rolling bearing is an important field that has received increasing attention in recent years. The main challenge for this task is the lack of labeled data. Existing works circumvent this problem with pseudo labels generated from labeled data. However, these pseudo labels are noisy even with consistency check or confidence-based filtering due to the minimal amount of training data. To solve this problem, a novel Label-level Anti-noise Fault Diagnosis Network (LAFD-Net) is proposed in this paper. Specifically, we propose an Online Asymptotic Label Updating (OALU) strategy which contains two updating stages: self-correction stage and cross-correction stage. The proposed OALU can stably and reliably generate new corrected pseudo labels, gradually replacing the old noisy ones. The LAFD-Net adopts a Student-Teacher architecture. For such a Student-Teacher model, we propose a Consistency Enhancement (CE) loss to strengthen the feature consistency between student and teacher networks, aiming to achieve more efficient use of plentiful unlabeled data via feature regularization. Finally, a series of experiments were conducted using the University of Cincinnati Intelligent Maintenance System (IMS) Center dataset and the Case Western Reserve University (CWRU) bearing dataset. The experimental results demonstrate that the proposed semi-supervised learning schemes outperformed existing state-of-the-art methods with the same percentage of labeled data samples.</description><subject>Bearing fault</subject><subject>Data models</subject><subject>Deep learning</subject><subject>Fault diagnosis</subject><subject>Feature extraction</subject><subject>learning with noisy data</subject><subject>Noise measurement</subject><subject>semi-supervised Learning</subject><subject>Sensors</subject><subject>student-teacher model</subject><subject>Training</subject><issn>1530-437X</issn><fulltext>true</fulltext><rsrctype>article</rsrctype><creationdate>2023</creationdate><recordtype>article</recordtype><sourceid>RIE</sourceid><recordid>eNqFjsuKwkAQAOegYHx8gOChfyDZmUxiyN58hUVCECLoTUbSai8xkelE8e9V2Pue6lAUlBBjJT2lZPy1zleZ50tfe9rXOg6jjnBUqKUb6GjfE33mXylVHIWRI3bpLFm6GTbfkKKxFVVneFBzgawmfsKGsS1qSM0RS4ZTbSHHK7l5e0N7J8YC5u_qEyWmLRtYkjlXNRMPRfdkSsbRHwdikqy2ix-XEPFws3Q19nlQ749gGgf6H_0C8FI_tg</recordid><startdate>20230110</startdate><enddate>20230110</enddate><creator>Jian, Yifan</creator><creator>Chen, Zhi</creator><creator>Lei, Yinjie</creator><creator>He, Zhengxi</creator><creator>Zhao, Yang</creator><creator>He, Liang</creator><creator>Luo, Wei</creator><creator>Chen, Xuekun</creator><general>IEEE</general><scope>97E</scope><scope>RIA</scope><scope>RIE</scope><orcidid>https://orcid.org/0000-0002-5820-9080</orcidid><orcidid>https://orcid.org/0000-0002-2683-384X</orcidid><orcidid>https://orcid.org/0000-0002-2683-384X</orcidid><orcidid>https://orcid.org/0000-0002-8211-3046</orcidid><orcidid>https://orcid.org/0000-0001-6856-3342</orcidid><orcidid>https://orcid.org/0000-0002-2683-384X</orcidid><orcidid>https://orcid.org/0000-0002-2683-384X</orcidid></search><sort><creationdate>20230110</creationdate><title>LAFD-Net: Learning with Noisy Pseudo Labels for Semi-Supervised Bearing Fault Diagnosis</title><author>Jian, Yifan ; Chen, Zhi ; Lei, Yinjie ; He, Zhengxi ; Zhao, Yang ; He, Liang ; Luo, Wei ; Chen, Xuekun</author></sort><facets><frbrtype>5</frbrtype><frbrgroupid>cdi_FETCH-ieee_primary_100146943</frbrgroupid><rsrctype>articles</rsrctype><prefilter>articles</prefilter><language>eng</language><creationdate>2023</creationdate><topic>Bearing fault</topic><topic>Data models</topic><topic>Deep learning</topic><topic>Fault diagnosis</topic><topic>Feature extraction</topic><topic>learning with noisy data</topic><topic>Noise measurement</topic><topic>semi-supervised Learning</topic><topic>Sensors</topic><topic>student-teacher model</topic><topic>Training</topic><toplevel>peer_reviewed</toplevel><toplevel>online_resources</toplevel><creatorcontrib>Jian, Yifan</creatorcontrib><creatorcontrib>Chen, Zhi</creatorcontrib><creatorcontrib>Lei, Yinjie</creatorcontrib><creatorcontrib>He, Zhengxi</creatorcontrib><creatorcontrib>Zhao, Yang</creatorcontrib><creatorcontrib>He, Liang</creatorcontrib><creatorcontrib>Luo, Wei</creatorcontrib><creatorcontrib>Chen, Xuekun</creatorcontrib><collection>IEEE All-Society Periodicals Package (ASPP) 2005-present</collection><collection>IEEE All-Society Periodicals Package (ASPP) 1998-Present</collection><collection>IEEE Electronic Library (IEL)</collection><jtitle>IEEE sensors journal</jtitle></facets><delivery><delcategory>Remote Search Resource</delcategory><fulltext>fulltext_linktorsrc</fulltext></delivery><addata><au>Jian, Yifan</au><au>Chen, Zhi</au><au>Lei, Yinjie</au><au>He, Zhengxi</au><au>Zhao, Yang</au><au>He, Liang</au><au>Luo, Wei</au><au>Chen, Xuekun</au><format>journal</format><genre>article</genre><ristype>JOUR</ristype><atitle>LAFD-Net: Learning with Noisy Pseudo Labels for Semi-Supervised Bearing Fault Diagnosis</atitle><jtitle>IEEE sensors journal</jtitle><stitle>JSEN</stitle><date>2023-01-10</date><risdate>2023</risdate><spage>1</spage><epage>1</epage><pages>1-1</pages><issn>1530-437X</issn><coden>ISJEAZ</coden><abstract>Fault Diagnosis for the rolling bearing is an important field that has received increasing attention in recent years. The main challenge for this task is the lack of labeled data. Existing works circumvent this problem with pseudo labels generated from labeled data. However, these pseudo labels are noisy even with consistency check or confidence-based filtering due to the minimal amount of training data. To solve this problem, a novel Label-level Anti-noise Fault Diagnosis Network (LAFD-Net) is proposed in this paper. Specifically, we propose an Online Asymptotic Label Updating (OALU) strategy which contains two updating stages: self-correction stage and cross-correction stage. The proposed OALU can stably and reliably generate new corrected pseudo labels, gradually replacing the old noisy ones. The LAFD-Net adopts a Student-Teacher architecture. For such a Student-Teacher model, we propose a Consistency Enhancement (CE) loss to strengthen the feature consistency between student and teacher networks, aiming to achieve more efficient use of plentiful unlabeled data via feature regularization. Finally, a series of experiments were conducted using the University of Cincinnati Intelligent Maintenance System (IMS) Center dataset and the Case Western Reserve University (CWRU) bearing dataset. The experimental results demonstrate that the proposed semi-supervised learning schemes outperformed existing state-of-the-art methods with the same percentage of labeled data samples.</abstract><pub>IEEE</pub><doi>10.1109/JSEN.2023.3233957</doi><orcidid>https://orcid.org/0000-0002-5820-9080</orcidid><orcidid>https://orcid.org/0000-0002-2683-384X</orcidid><orcidid>https://orcid.org/0000-0002-2683-384X</orcidid><orcidid>https://orcid.org/0000-0002-8211-3046</orcidid><orcidid>https://orcid.org/0000-0001-6856-3342</orcidid><orcidid>https://orcid.org/0000-0002-2683-384X</orcidid><orcidid>https://orcid.org/0000-0002-2683-384X</orcidid></addata></record> |
fulltext | fulltext_linktorsrc |
identifier | ISSN: 1530-437X |
ispartof | IEEE sensors journal, 2023-01, p.1-1 |
issn | 1530-437X |
language | eng |
recordid | cdi_ieee_primary_10014694 |
source | IEEE Electronic Library (IEL) |
subjects | Bearing fault Data models Deep learning Fault diagnosis Feature extraction learning with noisy data Noise measurement semi-supervised Learning Sensors student-teacher model Training |
title | LAFD-Net: Learning with Noisy Pseudo Labels for Semi-Supervised Bearing Fault Diagnosis |
url | https://sfx.bib-bvb.de/sfx_tum?ctx_ver=Z39.88-2004&ctx_enc=info:ofi/enc:UTF-8&ctx_tim=2024-12-23T09%3A15%3A31IST&url_ver=Z39.88-2004&url_ctx_fmt=infofi/fmt:kev:mtx:ctx&rfr_id=info:sid/primo.exlibrisgroup.com:primo3-Article-ieee_RIE&rft_val_fmt=info:ofi/fmt:kev:mtx:journal&rft.genre=article&rft.atitle=LAFD-Net:%20Learning%20with%20Noisy%20Pseudo%20Labels%20for%20Semi-Supervised%20Bearing%20Fault%20Diagnosis&rft.jtitle=IEEE%20sensors%20journal&rft.au=Jian,%20Yifan&rft.date=2023-01-10&rft.spage=1&rft.epage=1&rft.pages=1-1&rft.issn=1530-437X&rft.coden=ISJEAZ&rft_id=info:doi/10.1109/JSEN.2023.3233957&rft_dat=%3Cieee_RIE%3E10014694%3C/ieee_RIE%3E%3Curl%3E%3C/url%3E&disable_directlink=true&sfx.directlink=off&sfx.report_link=0&rft_id=info:oai/&rft_id=info:pmid/&rft_ieee_id=10014694&rfr_iscdi=true |