Jun Shi
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Jun Shi has a total of 30 co-authors. Here are the top co-authors by number of shared publications:
- Wanyong Liang
Zhengzhou University of Light Industry
5 shared publications - Weifeng Cao
Zhengzhou University of Light Industry
4 shared publications - Wanyong Liang
The School of Electrical and Information Engineering, Zhengzhou University of Light Industry, Zhengzhou, China
3 shared publications - Yong Jiang
College of Electric and Information Engineering, Zhengzhou University of Light Industry, Zhengzhou, People’s Republic of China
3 shared publications - Hechao Hou
Cutting Products Department, Zhengzhou GL Tech Co., Ltd, Zhengzhou, People’s Republic of China
2 shared publications - Jian Lan
Guangli Ruihong Electronic Technology Co., Ltd
2 shared publications
Latest Publications
- 1 June 2026
- Published by IOP Publishing in Engineering Research Express
- Vol. 8 (11) , 115409
Hybrid signal-neural processing: MA-Fourier dual-branch forecasting for efficient time series prediction
Time series forecasting plays a critical role in industrial prognostics and health management, where both prediction accuracy and computational efficiency are essential. In practical industrial scenarios, forecasting models are often deployed under limited computational resources and must respond rapidly to dynamic operating conditions. This imposes significant challenges in achieving reliable predictions without incurring excessive computational cost. To address this challenge, we propose LeftNet, a lightweight hybrid signal-neural forecasting model with a dual-branch architecture. Specifically, a trend modeling module is designed to capture smooth temporal evolution through moving average filtering and multilayer perceptrons, while a high-frequency pattern module leverages Fourier-based spectral filtering to model short-term fluctuations and transient dynamics. By explicitly decoupling temporal patterns at different frequency scales, the proposed architecture enables efficient and interpretable feature extraction. Extensive experiments on both an industrial dataset (DST) and a public benchmark (ETT) demonstrate that LeftNet achieves competitive predictive performance with reduced computational cost. In particular, the proposed method maintains low prediction errors while achieving faster inference compared to representative baseline models. These results highlight the effectiveness of LeftNet in achieving a favorable trade-off between prediction accuracy and computational efficiency, making it well suited for real-time and resource-constrained forecasting applications.
- 1 June 2026
- Published by IOP Publishing in Engineering Research Express
- Vol. 8 (11) , 115409
Hybrid signal-neural processing: MA-Fourier dual-branch forecasting for efficient time series prediction
Time series forecasting plays a critical role in industrial prognostics and health management, where both prediction accuracy and computational efficiency are essential. In practical industrial scenarios, forecasting models are often deployed under limited computational resources and must respond rapidly to dynamic operating conditions. This imposes significant challenges in achieving reliable predictions without incurring excessive computational cost. To address this challenge, we propose LeftNet, a lightweight hybrid signal-neural forecasting model with a dual-branch architecture. Specifically, a trend modeling module is designed to capture smooth temporal evolution through moving average filtering and multilayer perceptrons, while a high-frequency pattern module leverages Fourier-based spectral filtering to model short-term fluctuations and transient dynamics. By explicitly decoupling temporal patterns at different frequency scales, the proposed architecture enables efficient and interpretable feature extraction. Extensive experiments on both an industrial dataset (DST) and a public benchmark (ETT) demonstrate that LeftNet achieves competitive predictive performance with reduced computational cost. In particular, the proposed method maintains low prediction errors while achieving faster inference compared to representative baseline models. These results highlight the effectiveness of LeftNet in achieving a favorable trade-off between prediction accuracy and computational efficiency, making it well suited for real-time and resource-constrained forecasting applications.
- 1 April 2026
- Published by IOP Publishing in Engineering Research Express
- Vol. 8 (7) , 075407
Prediction of tool wear in dicing machine using GRU-Wavelet-Attention-LSTM
Tool monitoring and remaining useful life (RUL) prediction are crucial for cost reduction, efficiency improvement, and ensuring machining quality. Addressing two core challenges in dicing machine tool wear: intermittent data loss and strong non-stationary noise interference. This study developed an end-to-end deep learning model named GRU-Wavelet-Attention-LSTM (GWAL). The model’s front-end employs a dual-stream preprocessing architecture comprising a Gated Recurrent Unit (GRU) and a Differentiable Wavelet Deconvolution Layer (DWDL) based on soft thresholds. Through temporal reconstruction and gradient-friendly frequency domain filtering, this structure effectively extracts deep features representing tool health status. Building upon this foundation, the model further integrates a Self-Attention mechanism with a Long Short-Term Memory (LSTM) network for joint modeling. By explicitly computing intra-sequence correlations, it dynamically focuses on wear acceleration inflection points or transient impact events. Assigning interpretable weights enables precise establishment of long-term dependencies in tool wear. System validation was conducted using a full-lifecycle dataset from a dicing machine. Results demonstrate that the GWAL model achieves a 4.3% relative improvement in coefficient of determination (R 2 ) compared to the advanced LSTM-TCN-Self-Attention model. It achieved relative reductions of 40.91% and 38.46% in mean absolute error (MAE) and root mean square error (RMSE), respectively. The GWAL model features a compact structure with minimal parameters. It demonstrates exceptional robustness against dirty data and high-precision prediction capabilities. It holds significant potential for industrial applications.
- 1 April 2026
- Published by IOP Publishing in Engineering Research Express
- Vol. 8 (7) , 075407
Prediction of tool wear in dicing machine using GRU-Wavelet-Attention-LSTM
Tool monitoring and remaining useful life (RUL) prediction are crucial for cost reduction, efficiency improvement, and ensuring machining quality. Addressing two core challenges in dicing machine tool wear: intermittent data loss and strong non-stationary noise interference. This study developed an end-to-end deep learning model named GRU-Wavelet-Attention-LSTM (GWAL). The model’s front-end employs a dual-stream preprocessing architecture comprising a Gated Recurrent Unit (GRU) and a Differentiable Wavelet Deconvolution Layer (DWDL) based on soft thresholds. Through temporal reconstruction and gradient-friendly frequency domain filtering, this structure effectively extracts deep features representing tool health status. Building upon this foundation, the model further integrates a Self-Attention mechanism with a Long Short-Term Memory (LSTM) network for joint modeling. By explicitly computing intra-sequence correlations, it dynamically focuses on wear acceleration inflection points or transient impact events. Assigning interpretable weights enables precise establishment of long-term dependencies in tool wear. System validation was conducted using a full-lifecycle dataset from a dicing machine. Results demonstrate that the GWAL model achieves a 4.3% relative improvement in coefficient of determination (R 2 ) compared to the advanced LSTM-TCN-Self-Attention model. It achieved relative reductions of 40.91% and 38.46% in mean absolute error (MAE) and root mean square error (RMSE), respectively. The GWAL model features a compact structure with minimal parameters. It demonstrates exceptional robustness against dirty data and high-precision prediction capabilities. It holds significant potential for industrial applications.
- 12 August 2025
- Published by Springer Nature in The International Journal of Advanced Manufacturing Technology
- Vol. 139 (11-12) , 6107-6122
Wafer sub-pixel alignment based on multi-scale residual channel attention network and multi-strategy improved sparrow search algorithm
- 12 August 2025
- Published by Springer Nature in The International Journal of Advanced Manufacturing Technology
- Vol. 139 (11-12) , 6107-6122
Wafer sub-pixel alignment based on multi-scale residual channel attention network and multi-strategy improved sparrow search algorithm
- Open Access
DIFFUMA: High-Fidelity Spatio-Temporal Video Prediction via Dual-Path Mamba and Diffusion Enhancement
Spatio-temporal video prediction plays a pivotal role in critical domains, ranging from weather forecasting to industrial automation. However, in high-precision industrial scenarios such as semiconductor manufacturing, the absence of specialized benchmark datasets severely hampers research on modeling and predicting complex processes. To address this challenge, we make a twofold contribution.First, we construct and release the Chip Dicing Lane Dataset (CHDL), the first public temporal image dataset dedicated to the semiconductor wafer dicing process. Captured via an industrial-grade vision system, CHDL provides a much-needed and challenging benchmark for high-fidelity process modeling, defect detection, and digital twin development.Second, we propose DIFFUMA, an innovative dual-path prediction architecture specifically designed for such fine-grained dynamics. The model captures global long-range temporal context through a parallel Mamba module, while simultaneously leveraging a diffusion module, guided by temporal features, to restore and enhance fine-grained spatial details, effectively combating feature degradation. Experiments demonstrate that on our CHDL benchmark, DIFFUMA significantly outperforms existing methods, reducing the Mean Squared Error (MSE) by 39% and improving the Structural Similarity (SSIM) from 0.926 to a near-perfect 0.988. This superior performance also generalizes to natural phenomena datasets. Our work not only delivers a new state-of-the-art (SOTA) model but, more importantly, provides the community with an invaluable data resource to drive future research in industrial AI.
Open AccessDIFFUMA: High-Fidelity Spatio-Temporal Video Prediction via Dual-Path Mamba and Diffusion Enhancement
Spatio-temporal video prediction plays a pivotal role in critical domains, ranging from weather forecasting to industrial automation. However, in high-precision industrial scenarios such as semiconductor manufacturing, the absence of specialized benchmark datasets severely hampers research on modeling and predicting complex processes. To address this challenge, we make a twofold contribution.First, we construct and release the Chip Dicing Lane Dataset (CHDL), the first public temporal image dataset dedicated to the semiconductor wafer dicing process. Captured via an industrial-grade vision system, CHDL provides a much-needed and challenging benchmark for high-fidelity process modeling, defect detection, and digital twin development.Second, we propose DIFFUMA, an innovative dual-path prediction architecture specifically designed for such fine-grained dynamics. The model captures global long-range temporal context through a parallel Mamba module, while simultaneously leveraging a diffusion module, guided by temporal features, to restore and enhance fine-grained spatial details, effectively combating feature degradation. Experiments demonstrate that on our CHDL benchmark, DIFFUMA significantly outperforms existing methods, reducing the Mean Squared Error (MSE) by 39% and improving the Structural Similarity (SSIM) from 0.926 to a near-perfect 0.988. This superior performance also generalizes to natural phenomena datasets. Our work not only delivers a new state-of-the-art (SOTA) model but, more importantly, provides the community with an invaluable data resource to drive future research in industrial AI.
- 25 November 2024
- Published by Springer Nature in Microsystem Technologies
- Vol. 31 (6) , 1547-1560
Design of position controller for dicing saw chuck table based on improved LADRC technology
- 25 November 2024
- Published by Springer Nature in Microsystem Technologies
- Vol. 31 (6) , 1547-1560
Design of position controller for dicing saw chuck table based on improved LADRC technology
- Open Access
HASN: Hybrid Attention Separable Network for Efficient Image Super-resolution
Recently, lightweight methods for single image super-resolution (SISR) have gained significant popularity and achieved impressive performance due to limited hardware resources. These methods demonstrate that adopting residual feature distillation is an effective way to enhance performance. However, we find that using residual connections after each block increases the model's storage and computational cost. Therefore, to simplify the network structure and learn higher-level features and relationships between features, we use depthwise separable convolutions, fully connected layers, and activation functions as the basic feature extraction modules. This significantly reduces computational load and the number of parameters while maintaining strong feature extraction capabilities. To further enhance model performance, we propose the Hybrid Attention Separable Block (HASB), which combines channel attention and spatial attention, thus making use of their complementary advantages. Additionally, we use depthwise separable convolutions instead of standard convolutions, significantly reducing the computational load and the number of parameters while maintaining strong feature extraction capabilities. During the training phase, we also adopt a warm-start retraining strategy to exploit the potential of the model further. Extensive experiments demonstrate the effectiveness of our approach. Our method achieves a smaller model size and reduced computational complexity without compromising performance. Code can be available at https://github.com/nathan66666/HASN.git
Open AccessHASN: Hybrid Attention Separable Network for Efficient Image Super-resolution
Recently, lightweight methods for single image super-resolution (SISR) have gained significant popularity and achieved impressive performance due to limited hardware resources. These methods demonstrate that adopting residual feature distillation is an effective way to enhance performance. However, we find that using residual connections after each block increases the model's storage and computational cost. Therefore, to simplify the network structure and learn higher-level features and relationships between features, we use depthwise separable convolutions, fully connected layers, and activation functions as the basic feature extraction modules. This significantly reduces computational load and the number of parameters while maintaining strong feature extraction capabilities. To further enhance model performance, we propose the Hybrid Attention Separable Block (HASB), which combines channel attention and spatial attention, thus making use of their complementary advantages. Additionally, we use depthwise separable convolutions instead of standard convolutions, significantly reducing the computational load and the number of parameters while maintaining strong feature extraction capabilities. During the training phase, we also adopt a warm-start retraining strategy to exploit the potential of the model further. Extensive experiments demonstrate the effectiveness of our approach. Our method achieves a smaller model size and reduced computational complexity without compromising performance. Code can be available at https://github.com/nathan66666/HASN.git
- 28 September 2024
- Published by Springer Nature in Journal of Real-Time Image Processing
- Vol. 21 (5) , 1-10
RDB-YOLOv8n: Insulator defect detection based on improved lightweight YOLOv8n model
- 28 September 2024
- Published by Springer Nature in Journal of Real-Time Image Processing
- Vol. 21 (5) , 1-10
RDB-YOLOv8n: Insulator defect detection based on improved lightweight YOLOv8n model
- 3 September 2024
- Published by Springer Nature in The Visual Computer
- Vol. 41 (5) , 3423-3435
HASN: hybrid attention separable network for efficient image super-resolution
- 3 September 2024
- Published by Springer Nature in The Visual Computer
- Vol. 41 (5) , 3423-3435
HASN: hybrid attention separable network for efficient image super-resolution
- 6 April 2024
- Published by Springer Nature in International Journal of System Assurance Engineering and Management
- Vol. 15 (7) , 3104-3116
Optimization of servo accuracy of Y axis of dicing saw based on iterative learning control
- 6 April 2024
- Published by Springer Nature in International Journal of System Assurance Engineering and Management
- Vol. 15 (7) , 3104-3116
Optimization of servo accuracy of Y axis of dicing saw based on iterative learning control
- 8 November 2023
- Published by Springer Nature in The Journal of Supercomputing
- Vol. 80 (6) , 7483-7506
Chipping value prediction for dicing saw based on sparrow search algorithm and neural networks
- 8 November 2023
- Published by Springer Nature in The Journal of Supercomputing
- Vol. 80 (6) , 7483-7506
Chipping value prediction for dicing saw based on sparrow search algorithm and neural networks