Congyan Chen
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Congyan Chen has a total of 49 co-authors. Here are the top co-authors by number of shared publications:
- Dezheng Wang
School of Automation, Southeast University, Nanjing 210096, People’s Republic of China
8 shared publications - Shihua Li
School of Automation and Key Laboratory of Measurement and Control of Complex Systems of Engineering, Ministry of Education, Southeast University, Nanjing, China
8 shared publications - Jinya Su
School of Automation, the Key Laboratory of Measurement and Control of CSE, Ministry of Education, and the Institute of Intelligent Unmanned Systems, Southeast University, Nanjing, China
3 shared publications - Fuyuan Xiang
School of Automation, Southeast University, Nanjing 210096, China
2 shared publications - Hongzhi Yin
The University of Queensland
2 shared publications - Ke Yu
School of Automation, Key Laboratory of Measurement and Control of CSE, Ministry of Education, Institute of Intelligent Unmanned Systems, Southeast University, Nanjing, China
2 shared publications
Latest Publications
- Open Access
- 8 August 2026
- Published by Association for Computing Machinery (ACM)
- p. 4800-4811
LEFT: Learnable Fusion of Tri-view Tokens for Unsupervised Time Series Anomaly Detection
- 8 August 2026
- Published by Association for Computing Machinery (ACM)
- p. 4800-4811
Open AccessLEFT: Learnable Fusion of Tri-view Tokens for Unsupervised Time Series Anomaly Detection
- 24 May 2026
- Published by Springer Nature in Circuits, Systems, and Signal Processing
- Vol. 45 (12) , 10231-10251
A Frequency-Independent Iterative DFT Interpolation Method for Robust Sinusoidal Frequency Estimation
- 24 May 2026
- Published by Springer Nature in Circuits, Systems, and Signal Processing
- Vol. 45 (12) , 10231-10251
A Frequency-Independent Iterative DFT Interpolation Method for Robust Sinusoidal Frequency Estimation
- 3 March 2026
- Published by Springer Nature in Journal of Real-Time Image Processing
- Vol. 23 (2) , 68
A novel real-time image defect detection network for photovoltaic modules
- 3 March 2026
- Published by Springer Nature in Journal of Real-Time Image Processing
- Vol. 23 (2) , 68
A novel real-time image defect detection network for photovoltaic modules
- 6 February 2026
- Published by Springer Nature in Journal of Real-Time Image Processing
- Vol. 23 (2) , 59
Tiny-YOLOv8: an embedded deployment approach engineering realization for electroluminescence image detection
- 6 February 2026
- Published by Springer Nature in Journal of Real-Time Image Processing
- Vol. 23 (2) , 59
Tiny-YOLOv8: an embedded deployment approach engineering realization for electroluminescence image detection
- 4 February 2026
- Published by Elsevier in Measurement
- Vol. 268
MACo-Net: a multi-domain adaptive collaborative network for few-shot fault diagnosis
- 4 February 2026
- Published by Elsevier in Measurement
- Vol. 268
MACo-Net: a multi-domain adaptive collaborative network for few-shot fault diagnosis
- 1 August 2025
- Published by IOP Publishing in Measurement Science and Technology
- Vol. 36 (8) , 086111
A sliding window Transformer model for improved local features in mechanical equipment fault detection
Mechanical equipment fault detection plays a vital role in modern industry and manufacturing. Many existing models, however, either favor global features or focus on local features during features extraction, limiting the detection accuracy and generalization capability of the models. This paper proposes a sliding window Transformer (SWT) model for enhancing local multi-domain features extraction from acoustic signals of mechanical equipment. Compared with the traditional Transformer, the SWT model applies the sliding window principle to mask the attention distribution matrix, retaining local features. These local features are subsequently fused with the global features extracted by the attention mechanism. This approach enables the model to preserve its capacity for capturing global features while enhancing local details, thus achieving efficient features representation. Experimental results indicate that SWT achieves higher fault detection accuracy than Transformer under various signal-to-noise ratio conditions specially with a 1.9% improvement on the MIMII-Gearbox dataset at −2 dB. In addition, the method keeps better classification performance under industrial noise conditions compared with other models. Comparative experimental results validate the effectiveness and feasibility of the proposed method for practical applications.
- 1 August 2025
- Published by IOP Publishing in Measurement Science and Technology
- Vol. 36 (8) , 086111
A sliding window Transformer model for improved local features in mechanical equipment fault detection
Mechanical equipment fault detection plays a vital role in modern industry and manufacturing. Many existing models, however, either favor global features or focus on local features during features extraction, limiting the detection accuracy and generalization capability of the models. This paper proposes a sliding window Transformer (SWT) model for enhancing local multi-domain features extraction from acoustic signals of mechanical equipment. Compared with the traditional Transformer, the SWT model applies the sliding window principle to mask the attention distribution matrix, retaining local features. These local features are subsequently fused with the global features extracted by the attention mechanism. This approach enables the model to preserve its capacity for capturing global features while enhancing local details, thus achieving efficient features representation. Experimental results indicate that SWT achieves higher fault detection accuracy than Transformer under various signal-to-noise ratio conditions specially with a 1.9% improvement on the MIMII-Gearbox dataset at −2 dB. In addition, the method keeps better classification performance under industrial noise conditions compared with other models. Comparative experimental results validate the effectiveness and feasibility of the proposed method for practical applications.
- 23 April 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Sensors Journal
- Vol. 25 (11) , 20024-20036
BeltLineNet: A Shape-Prior-Guided Lightweight Network for Real-Time Deviation Detection in Circular Pipe Conveyors
Belt deviation in circular pipe conveyor systems could lead to material spillage, environmental contamination, reduced efficiency, and accelerated belt wear. Real-time belt deviation detection is crucial for ensuring safe and efficient operation. However, existing methods, primarily developed for flat-belt systems, heavily rely on distinct belt-background semantics, making them unsuitable for circular conveyors where indistinct target features, motion blur, and occlusions pose significant challenges. To address these issues, we reformulate the deviation detection task as an object detection problem, proposing BeltLineNet, a lightweight, real-time deviation detection network specifically designed for circular pipe conveyors. Our method integrates explicit feature learning with a coarse-to-fine global feature fusion mechanism, enhancing belt line representation under complex conditions. Additionally, a shape-prior loss strategy is introduced to improve supervision during training, ensuring more accurate detection of elongated targets. The model is also optimized for real-time deployment through layer-adaptive pruning, achieving a balance between accuracy and computational efficiency. Extensive comparative experiments involving 8 state-of-the-art (SOTA) feature extraction networks and 6 feature enhancement strategies on a self-collected dataset under industrial conditions demonstrate that BeltLineNet surpasses SOTA feature extraction and fusion networks, improving average precision (AP) by 8.23% before pruning and 6.33% after pruning.
- 23 April 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Sensors Journal
- Vol. 25 (11) , 20024-20036
BeltLineNet: A Shape-Prior-Guided Lightweight Network for Real-Time Deviation Detection in Circular Pipe Conveyors
Belt deviation in circular pipe conveyor systems could lead to material spillage, environmental contamination, reduced efficiency, and accelerated belt wear. Real-time belt deviation detection is crucial for ensuring safe and efficient operation. However, existing methods, primarily developed for flat-belt systems, heavily rely on distinct belt-background semantics, making them unsuitable for circular conveyors where indistinct target features, motion blur, and occlusions pose significant challenges. To address these issues, we reformulate the deviation detection task as an object detection problem, proposing BeltLineNet, a lightweight, real-time deviation detection network specifically designed for circular pipe conveyors. Our method integrates explicit feature learning with a coarse-to-fine global feature fusion mechanism, enhancing belt line representation under complex conditions. Additionally, a shape-prior loss strategy is introduced to improve supervision during training, ensuring more accurate detection of elongated targets. The model is also optimized for real-time deployment through layer-adaptive pruning, achieving a balance between accuracy and computational efficiency. Extensive comparative experiments involving 8 state-of-the-art (SOTA) feature extraction networks and 6 feature enhancement strategies on a self-collected dataset under industrial conditions demonstrate that BeltLineNet surpasses SOTA feature extraction and fusion networks, improving average precision (AP) by 8.23% before pruning and 6.33% after pruning.