Yi Chen
Research Subjects
Yi Chen has a total of 10 co-authors. Here are the top co-authors by number of shared publications:
Bojun Long
Xi'an Jiaotong-Liverpool University,School of Intelligent Manufacturing Ecosystem, XJTLU Entrepreneur College (Taicang),Suzhou,China
1 shared publication- Jiabao An
Industrial Intelligent Agent Research Center, School of Intelligent Manufacturing Ecosystem, XJTLU Entrepreneur College (Taicang), Xi’an Jiaotong-Liverpool University, Suzhou 215123, China
1 shared publication - Mengjie Huang
Design School, Xian Jiaotong-Liverpool University, Suzhou, China
1 shared publication - Rui Yang
School of Advanced Technology, Xi’an Jiaotong-Liverpool University, Suzhou 215123, China
1 shared publication Xiaohui Liu
Department of Computer Science, Brunel University London, Uxbridge, Middlesex, U.K
1 shared publication- Xiongfei Cai
Department of Electronic Engineering, Tsinghua University, Beijing, China
1 shared publication
Latest Publications
- 9 September 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Instrumentation and Measurement
- Vol. 74 (00189456) , 1-11
Defect Detection in Patterned Wafers Using Dynamic Local Adaptive Sensitivity Adjustment
In the defect detection of patterned wafers, optical images captured by inspection systems are affected by various noises, resulting in low signal-to-noise ratios in the obtained images, making the detection task extremely challenging. Especially at the nanoscale pattern dimensions, optical image differences formed by different wafer patterns are significant. Designing an automatic algorithm to achieve good detection results for various defect types and wafer patterns is difficult. This paper presents a novel dynamic local adaptive (DLA) sensitivity adjustment framework, which represents the first application of context-aware local region generation and pattern-based local threshold adjustment in semiconductor defect detection. DLA introduces a dual adaptive mechanism—dynamic window sizing and noise-aware threshold adjustment—that can be seamlessly integrated with existing thresholding frameworks, achieving state-of-the-art performance without hardware upgrades. DLA is validated on the data of a standard wafer designed by SIOK Technology Co. Ltd. to demonstrate the proposed method’s effectiveness. Additionally, this method is compared with other well-known methods in defect detection. Experimental results show that the proposed method is adaptable, can achieve better detection results, and lower false nuisance rates for various defect types.
- 9 September 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Instrumentation and Measurement
- Vol. 74 (00189456) , 1-11
Defect Detection in Patterned Wafers Using Dynamic Local Adaptive Sensitivity Adjustment
In the defect detection of patterned wafers, optical images captured by inspection systems are affected by various noises, resulting in low signal-to-noise ratios in the obtained images, making the detection task extremely challenging. Especially at the nanoscale pattern dimensions, optical image differences formed by different wafer patterns are significant. Designing an automatic algorithm to achieve good detection results for various defect types and wafer patterns is difficult. This paper presents a novel dynamic local adaptive (DLA) sensitivity adjustment framework, which represents the first application of context-aware local region generation and pattern-based local threshold adjustment in semiconductor defect detection. DLA introduces a dual adaptive mechanism—dynamic window sizing and noise-aware threshold adjustment—that can be seamlessly integrated with existing thresholding frameworks, achieving state-of-the-art performance without hardware upgrades. DLA is validated on the data of a standard wafer designed by SIOK Technology Co. Ltd. to demonstrate the proposed method’s effectiveness. Additionally, this method is compared with other well-known methods in defect detection. Experimental results show that the proposed method is adaptable, can achieve better detection results, and lower false nuisance rates for various defect types.
- 28 August 2023
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 1-10
Neural Radiance Fields Assist in Object Grab for Digital Twins
The Fourth Industrial Revolution, a transformative milestone in modern history, has introduced significant enhancements to traditional manufacturing industries and given rise to novel manufacturing systems. Industry 4.0 technology, which underpins this transformation, is characterized by the integration of digital twin technology—an essential component for bridging physical and digital information. However, current realtime digital twin systems predominantly focus on conveying mechanical information while largely neglecting macroscopic visual information observable by the naked eye. This paper proposes a method to augment digital twin technology by leveraging neural radiance fields, thereby facilitating the observation of dynamic non-mechanical information in real-world settings.
- 28 August 2023
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 1-10
Neural Radiance Fields Assist in Object Grab for Digital Twins
The Fourth Industrial Revolution, a transformative milestone in modern history, has introduced significant enhancements to traditional manufacturing industries and given rise to novel manufacturing systems. Industry 4.0 technology, which underpins this transformation, is characterized by the integration of digital twin technology—an essential component for bridging physical and digital information. However, current realtime digital twin systems predominantly focus on conveying mechanical information while largely neglecting macroscopic visual information observable by the naked eye. This paper proposes a method to augment digital twin technology by leveraging neural radiance fields, thereby facilitating the observation of dynamic non-mechanical information in real-world settings.
- Open Access
- 18 July 2022
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Neural Systems and Rehabilitation Engineering
- Vol. 30 (15344320) , 1992-2002
Single-Source to Single-Target Cross-Subject Motor Imagery Classification Based on Multisubdomain Adaptation Network
In the electroencephalography (EEG) based cross-subject motor imagery (MI) classification task, the device and subject problems can cause the time-related data distribution shift problem. In a single-source to single-target (STS) MI classification task, such a shift problem will certainly provoke an increase in the overall data distribution difference between the source and target domains, giving rise to poor classification accuracy. In this paper, a novel multi-subdomain adaptation method (MSDAN) is proposed to solve the shift problem and improve the classification accuracy of the traditional approaches. In the proposed MSDAN, the adaptation losses in both class-related and time-related subdomains (that are divided by different data labels and session labels) are obtained by measuring the distribution differences between the source and target subdomains. Then, the adaptation and classification losses in the loss function of MSDAN are minimized concurrently. To illustrate the application value of the proposed method, our method is applied to solve the STS MI classification task about data analysis with respect to the brain-computer interface (BCI) competition III-IVa dataset. The resultant experiment results demonstrate that compared with other well-known domain adaptation and deep learning methods, the proposed method is capable of solving the time-related data distribution problem at higher classification accuracy.
- 18 July 2022
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Neural Systems and Rehabilitation Engineering
- Vol. 30 (15344320) , 1992-2002
Open AccessSingle-Source to Single-Target Cross-Subject Motor Imagery Classification Based on Multisubdomain Adaptation Network
In the electroencephalography (EEG) based cross-subject motor imagery (MI) classification task, the device and subject problems can cause the time-related data distribution shift problem. In a single-source to single-target (STS) MI classification task, such a shift problem will certainly provoke an increase in the overall data distribution difference between the source and target domains, giving rise to poor classification accuracy. In this paper, a novel multi-subdomain adaptation method (MSDAN) is proposed to solve the shift problem and improve the classification accuracy of the traditional approaches. In the proposed MSDAN, the adaptation losses in both class-related and time-related subdomains (that are divided by different data labels and session labels) are obtained by measuring the distribution differences between the source and target subdomains. Then, the adaptation and classification losses in the loss function of MSDAN are minimized concurrently. To illustrate the application value of the proposed method, our method is applied to solve the STS MI classification task about data analysis with respect to the brain-computer interface (BCI) competition III-IVa dataset. The resultant experiment results demonstrate that compared with other well-known domain adaptation and deep learning methods, the proposed method is capable of solving the time-related data distribution problem at higher classification accuracy.