As samples of steel defects are industrially limited, it is challenging for most deep learning methods that rely on ample labeled data to identify steel surface defects. Recently, contrastive learning has achieved good performance in natural image classification tasks with few labeled samples, yet two obstacles prevent its effective application to steel surface defect images. One is that due to the presence of inter-class and intra-class similar samples in steel surface defect, the fixed contrast strength in contrastive learning will destroy the potential semantic information of defect samples. Another is that contrastive learning requires a large amount of unlabeled data, whereas steel surface defect samples are insufficient. To overcome the above-mentioned problems, a novel framework named flexible and diverse contrastive learning (FDCL) is proposed. This framework consists of two parts, flexible contrast (FiCo) and diverse generative adversarial networks (DGANs). Diverse images generated by the DGAN and real images are fed into FiCo for representation learning. In the FiCo, the contrast strength among samples is flexibly adjusted by the proposed variable temperature discrimination and feature reconstruction (FR). In addition, the output features (OF) of FiCo will be used as input to the DGAN generator to improve image quality, thus further facilitating representation learning. The proposed FDCL is implemented on four standard steel surface defect data sets, and the experimental results demonstrated that it achieves superior performance over state-of-the-art methods. Our code is available at: https://github.com-/jiacongc/FDCL .

Flexible and Diverse Contrastive Learning for Steel Surface Defect Recognition with Few Labeled Samples / Y. Xu, J. Chen, Y. Liang, Y. Zhai, Z. Ying, W. Zhou, A. Genovese, V. Piuri, F. Scotti. - In: IEEE TRANSACTIONS ON INSTRUMENTATION AND MEASUREMENT. - ISSN 0018-9456. - 72:(2023), pp. 2516014.1-2516014.14. [10.1109/TIM.2023.3249221]

Flexible and Diverse Contrastive Learning for Steel Surface Defect Recognition with Few Labeled Samples

A. Genovese;V. Piuri
Penultimo
;
F. Scotti
Ultimo
2023

Abstract

As samples of steel defects are industrially limited, it is challenging for most deep learning methods that rely on ample labeled data to identify steel surface defects. Recently, contrastive learning has achieved good performance in natural image classification tasks with few labeled samples, yet two obstacles prevent its effective application to steel surface defect images. One is that due to the presence of inter-class and intra-class similar samples in steel surface defect, the fixed contrast strength in contrastive learning will destroy the potential semantic information of defect samples. Another is that contrastive learning requires a large amount of unlabeled data, whereas steel surface defect samples are insufficient. To overcome the above-mentioned problems, a novel framework named flexible and diverse contrastive learning (FDCL) is proposed. This framework consists of two parts, flexible contrast (FiCo) and diverse generative adversarial networks (DGANs). Diverse images generated by the DGAN and real images are fed into FiCo for representation learning. In the FiCo, the contrast strength among samples is flexibly adjusted by the proposed variable temperature discrimination and feature reconstruction (FR). In addition, the output features (OF) of FiCo will be used as input to the DGAN generator to improve image quality, thus further facilitating representation learning. The proposed FDCL is implemented on four standard steel surface defect data sets, and the experimental results demonstrated that it achieves superior performance over state-of-the-art methods. Our code is available at: https://github.com-/jiacongc/FDCL .
Contrastive learning; diverse generative adversarial network (DGAN); feature reconstruction (FR); flexible contrast (FiCo); steel surface defect recognition; variable temperature discrimination
Settore INF/01 - Informatica
Settore ING-INF/05 - Sistemi di Elaborazione delle Informazioni
2023
Article (author)
File in questo prodotto:
File Dimensione Formato  
tim23.pdf

accesso aperto

Tipologia: Post-print, accepted manuscript ecc. (versione accettata dall'editore)
Dimensione 7.67 MB
Formato Adobe PDF
7.67 MB Adobe PDF Visualizza/Apri
Flexible_and_Diverse_Contrastive_Learning_for_Steel_Surface_Defect_Recognition_.pdf

accesso riservato

Tipologia: Publisher's version/PDF
Dimensione 6.81 MB
Formato Adobe PDF
6.81 MB Adobe PDF   Visualizza/Apri   Richiedi una copia
Pubblicazioni consigliate

I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.

Utilizza questo identificativo per citare o creare un link a questo documento: https://hdl.handle.net/2434/976788
Citazioni
  • ???jsp.display-item.citation.pmc??? ND
  • Scopus 3
  • ???jsp.display-item.citation.isi??? 2
social impact