Pengli Zhu
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Previous Organizations / Institutes
The organizations / institutes from the scholar's other articles in the past ten years
Pengli Zhu has a total of 30 co-authors. Here are the top co-authors by number of shared publications:
- Yancheng Liu
Marine Engineering College, Dalian Maritime University, Dalian 116026, China
6 shared publications - Anqi Qiu
Hong Kong Polytechnic University
4 shared publications - Siyuan Liu
Dalian Maritime University, Marine Engineering College, Dalian, China
4 shared publications - Haohao Guo
School of Marine Engineering, Dalian Maritime University, Dalian, China
3 shared publications - Nanguang Chen
Department of Biomedical Engineering, National University of Singapore, Singapore
3 shared publications - Xuzhou Zhuang
Marine Engineering College, Dalian Maritime University, Dalian, China
3 shared publications
Latest Publications
- 7 October 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Medical Imaging
- Vol. 45 (3) , 1167-1178
Q-Space Guided Multi-Modal Translation Network for Diffusion-Weighted Image Synthesis
Diffusion-weighted imaging (DWI) enables non-invasive characterization of tissue microstructure, yet acquiring densely sampled q-space data remains time-consuming and impractical in many clinical settings. Existing deep learning methods are typically constrained by fixed q-space sampling, limiting their adaptability to variable sampling scenarios. In this paper, we propose a Q-space Guided Multi-Modal Translation Network (Q-MMTN) for synthesizing multi-shell, high-angular resolution DWI (MS-HARDI) from flexible q-space sampling, leveraging commonly acquired structural data (e.g., T1- and T2-weighted MRI). Q-MMTN integrates the hybrid encoder and multi-modal attention fusion mechanism to effectively extract both local and global complementary information from multiple modalities. This design enhances feature representation and, together with a flexible q-space-aware embedding, enables dynamic modulation of internal features without relying on fixed sampling schemes. Additionally, we introduce a set of task-specific constraints, including adversarial, reconstruction, and anatomical consistency losses, which jointly enforce anatomical fidelity and signal realism. These constraints guide Q-MMTN to accurately capture the intrinsic and nonlinear relationships between directional DWI signals and q-space information. Extensive experiments across four lifespan datasets of children, adolescents, young and older adults demonstrate that Q-MMTN outperforms existing methods, including 1D-qDL, 2D-qDL, MESC-SD, and Q-GAN in estimating parameter maps and fiber tracts with fine-grained anatomical details. Notably, its ability to accommodate flexible q-space sampling highlights its potential as a promising toolkit for clinical and research applications. Our code is available at https://github.com/Idea89560041/Q-MMTN.
- 7 October 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Medical Imaging
- Vol. 45 (3) , 1167-1178
Q-Space Guided Multi-Modal Translation Network for Diffusion-Weighted Image Synthesis
Diffusion-weighted imaging (DWI) enables non-invasive characterization of tissue microstructure, yet acquiring densely sampled q-space data remains time-consuming and impractical in many clinical settings. Existing deep learning methods are typically constrained by fixed q-space sampling, limiting their adaptability to variable sampling scenarios. In this paper, we propose a Q-space Guided Multi-Modal Translation Network (Q-MMTN) for synthesizing multi-shell, high-angular resolution DWI (MS-HARDI) from flexible q-space sampling, leveraging commonly acquired structural data (e.g., T1- and T2-weighted MRI). Q-MMTN integrates the hybrid encoder and multi-modal attention fusion mechanism to effectively extract both local and global complementary information from multiple modalities. This design enhances feature representation and, together with a flexible q-space-aware embedding, enables dynamic modulation of internal features without relying on fixed sampling schemes. Additionally, we introduce a set of task-specific constraints, including adversarial, reconstruction, and anatomical consistency losses, which jointly enforce anatomical fidelity and signal realism. These constraints guide Q-MMTN to accurately capture the intrinsic and nonlinear relationships between directional DWI signals and q-space information. Extensive experiments across four lifespan datasets of children, adolescents, young and older adults demonstrate that Q-MMTN outperforms existing methods, including 1D-qDL, 2D-qDL, MESC-SD, and Q-GAN in estimating parameter maps and fiber tracts with fine-grained anatomical details. Notably, its ability to accommodate flexible q-space sampling highlights its potential as a promising toolkit for clinical and research applications. Our code is available at https://github.com/Idea89560041/Q-MMTN.
- Open Access
- 31 July 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Pattern Analysis and Machine Intelligence
- Vol. 47 (12) , 11022-11039
HL-HGAT: Heterogeneous Graph Attention Network via Hodge-Laplacian Operator
Graph neural networks (GNNs) have proven effective in capturing relationships among nodes in a graph. This study introduces a novel perspective by considering a graph as a simplicial complex, encompassing nodes, edges, triangles, and$k$ -simplices, enabling the definition of graph-structured data on any$k$ -simplex. We design a novel Hodge-Laplacian heterogeneous graph attention network (HL-HGAT) to learn heterogeneous signal representations across$k$ -simplices. The HL-HGAT incorporates three key components: HL convolutional filters (HL-filters), simplicial projection (SP), and simplicial attention pooling (SAP) operators, applied to$k$ -simplices. HL-filters leverage the unique topology of$k$ -simplices encoded by the Hodge-Laplacian (HL) operator, operating within the spectral domain of the$k$ -th HL operator. To address computation challenges, we introduce a polynomial approximation for HL-filters, exhibiting spatial localization properties. Additionally, we propose a pooling operator to coarsen$k$ -simplices, combining features through simplicial attention mechanisms of self-attention and cross-attention via transformers and SP operators, capturing topological interconnections across multiple dimensions of simplices. The HL-HGAT is comprehensively evaluated across diverse graph applications, including NP-hard problems, graph multi-label and classification challenges, and graph regression tasks in logistics, computer vision, biology, chemistry, and neuroscience. The results demonstrate the model’s efficacy and versatility in handling a wide range of graph-based scenarios.- 31 July 2025
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Pattern Analysis and Machine Intelligence
- Vol. 47 (12) , 11022-11039
Open AccessHL-HGAT: Heterogeneous Graph Attention Network via Hodge-Laplacian Operator
Graph neural networks (GNNs) have proven effective in capturing relationships among nodes in a graph. This study introduces a novel perspective by considering a graph as a simplicial complex, encompassing nodes, edges, triangles, and$k$ -simplices, enabling the definition of graph-structured data on any$k$ -simplex. We design a novel Hodge-Laplacian heterogeneous graph attention network (HL-HGAT) to learn heterogeneous signal representations across$k$ -simplices. The HL-HGAT incorporates three key components: HL convolutional filters (HL-filters), simplicial projection (SP), and simplicial attention pooling (SAP) operators, applied to$k$ -simplices. HL-filters leverage the unique topology of$k$ -simplices encoded by the Hodge-Laplacian (HL) operator, operating within the spectral domain of the$k$ -th HL operator. To address computation challenges, we introduce a polynomial approximation for HL-filters, exhibiting spatial localization properties. Additionally, we propose a pooling operator to coarsen$k$ -simplices, combining features through simplicial attention mechanisms of self-attention and cross-attention via transformers and SP operators, capturing topological interconnections across multiple dimensions of simplices. The HL-HGAT is comprehensively evaluated across diverse graph applications, including NP-hard problems, graph multi-label and classification challenges, and graph regression tasks in logistics, computer vision, biology, chemistry, and neuroscience. The results demonstrate the model’s efficacy and versatility in handling a wide range of graph-based scenarios. - Open Access
- 21 April 2025
- Published by Elsevier in Medical Image Analysis
- Vol. 103, 103579
Cycle-conditional diffusion model for noise correction of diffusion-weighted images using unpaired data
- 21 April 2025
- Published by Elsevier in Medical Image Analysis
- Vol. 103, 103579
Open AccessCycle-conditional diffusion model for noise correction of diffusion-weighted images using unpaired data
- Open Access
Advancing Graph Neural Networks with HL-HGAT: A Hodge-Laplacian and Attention Mechanism Approach for Heterogeneous Graph-Structured Data
Graph neural networks (GNNs) have proven effective in capturing relationships among nodes in a graph. This study introduces a novel perspective by considering a graph as a simplicial complex, encompassing nodes, edges, triangles, and $k$-simplices, enabling the definition of graph-structured data on any $k$-simplices. Our contribution is the Hodge-Laplacian heterogeneous graph attention network (HL-HGAT), designed to learn heterogeneous signal representations across $k$-simplices. The HL-HGAT incorporates three key components: HL convolutional filters (HL-filters), simplicial projection (SP), and simplicial attention pooling (SAP) operators, applied to $k$-simplices. HL-filters leverage the unique topology of $k$-simplices encoded by the Hodge-Laplacian (HL) operator, operating within the spectral domain of the $k$-th HL operator. To address computation challenges, we introduce a polynomial approximation for HL-filters, exhibiting spatial localization properties. Additionally, we propose a pooling operator to coarsen $k$-simplices, combining features through simplicial attention mechanisms of self-attention and cross-attention via transformers and SP operators, capturing topological interconnections across multiple dimensions of simplices. The HL-HGAT is comprehensively evaluated across diverse graph applications, including NP-hard problems, graph multi-label and classification challenges, and graph regression tasks in logistics, computer vision, biology, chemistry, and neuroscience. The results demonstrate the model's efficacy and versatility in handling a wide range of graph-based scenarios.
Open AccessAdvancing Graph Neural Networks with HL-HGAT: A Hodge-Laplacian and Attention Mechanism Approach for Heterogeneous Graph-Structured Data
Graph neural networks (GNNs) have proven effective in capturing relationships among nodes in a graph. This study introduces a novel perspective by considering a graph as a simplicial complex, encompassing nodes, edges, triangles, and $k$-simplices, enabling the definition of graph-structured data on any $k$-simplices. Our contribution is the Hodge-Laplacian heterogeneous graph attention network (HL-HGAT), designed to learn heterogeneous signal representations across $k$-simplices. The HL-HGAT incorporates three key components: HL convolutional filters (HL-filters), simplicial projection (SP), and simplicial attention pooling (SAP) operators, applied to $k$-simplices. HL-filters leverage the unique topology of $k$-simplices encoded by the Hodge-Laplacian (HL) operator, operating within the spectral domain of the $k$-th HL operator. To address computation challenges, we introduce a polynomial approximation for HL-filters, exhibiting spatial localization properties. Additionally, we propose a pooling operator to coarsen $k$-simplices, combining features through simplicial attention mechanisms of self-attention and cross-attention via transformers and SP operators, capturing topological interconnections across multiple dimensions of simplices. The HL-HGAT is comprehensively evaluated across diverse graph applications, including NP-hard problems, graph multi-label and classification challenges, and graph regression tasks in logistics, computer vision, biology, chemistry, and neuroscience. The results demonstrate the model's efficacy and versatility in handling a wide range of graph-based scenarios.
- 6 November 2023
- Published by Springer Nature in Applied Intelligence
- Vol. 53 (24) , 29974-29990
METER: Multi-task efficient transformer for no-reference image quality assessment
- 6 November 2023
- Published by Springer Nature in Applied Intelligence
- Vol. 53 (24) , 29974-29990
METER: Multi-task efficient transformer for no-reference image quality assessment
- 10 October 2023
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Journal of Oceanic Engineering
- Vol. 49 (1) , 48-65
Unsupervised Multiple Representation Disentanglement Framework for Improved Underwater Visual Perception
Visual perception is a key component for many underwater applications. However, visual imaging data typically suffer from different types of distortion, which seriously affect their security and reliability. Underwater image enhancement (UIE) is a crucial step to improve the perceptual quality of the environment. Recent deep learning techniques for UIE are predominantly supervised in nature and, hence, require imaging data of the same scene to be acquired in distortion-corrupted and distortion-free environments, which is impractical, especially when a sufficient sample size is needed for reliable network training. As a remedy, we propose an unsupervised multiple representation disentanglement (UMRD) framework for improved underwater visual perception. UMRD employs a unified multidomain unsupervised cycle-consistent adversarial translation to simultaneously disentangle the domain label indicating image type, content features reflecting structural semantics, and style information reflecting appearance from all-domain underwater images. The disentangled content feature and target style information are then utilized to generate the target distortion-free image. Furthermore, we present the underwater multidistortion image enhancement data set, a large-scale collection of 19 K underwater images (of poor and good quality) that can be used for improved underwater visual perception (e.g., decolorization, deblurring, and denoising). Experimental results demonstrate that UMRD, trained using unpaired data, exhibits prominent capabilities of image quality enhancement and performance improvement for subsequent processing tasks, i.e., object detection and semantic segmentation.
- 10 October 2023
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Journal of Oceanic Engineering
- Vol. 49 (1) , 48-65
Unsupervised Multiple Representation Disentanglement Framework for Improved Underwater Visual Perception
Visual perception is a key component for many underwater applications. However, visual imaging data typically suffer from different types of distortion, which seriously affect their security and reliability. Underwater image enhancement (UIE) is a crucial step to improve the perceptual quality of the environment. Recent deep learning techniques for UIE are predominantly supervised in nature and, hence, require imaging data of the same scene to be acquired in distortion-corrupted and distortion-free environments, which is impractical, especially when a sufficient sample size is needed for reliable network training. As a remedy, we propose an unsupervised multiple representation disentanglement (UMRD) framework for improved underwater visual perception. UMRD employs a unified multidomain unsupervised cycle-consistent adversarial translation to simultaneously disentangle the domain label indicating image type, content features reflecting structural semantics, and style information reflecting appearance from all-domain underwater images. The disentangled content feature and target style information are then utilized to generate the target distortion-free image. Furthermore, we present the underwater multidistortion image enhancement data set, a large-scale collection of 19 K underwater images (of poor and good quality) that can be used for improved underwater visual perception (e.g., decolorization, deblurring, and denoising). Experimental results demonstrate that UMRD, trained using unpaired data, exhibits prominent capabilities of image quality enhancement and performance improvement for subsequent processing tasks, i.e., object detection and semantic segmentation.
- 24 August 2023
- Published by Elsevier in Engineering Applications of Artificial Intelligence
- Vol. 126, 106866
Unsupervised underwater image enhancement via content-style representation disentanglement
- 24 August 2023
- Published by Elsevier in Engineering Applications of Artificial Intelligence
- Vol. 126, 106866
Unsupervised underwater image enhancement via content-style representation disentanglement
- 24 May 2022
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Vehicular Technology
- Vol. 71 (8) , 8237-8250
Autonomous Reinforcement Control of Visual Underwater Vehicles: Real-Time Experiments Using Computer Vision
Swift decision-making based on visual environment perception is crucial for autonomous control of visual underwater vehicles (VUVs) during underwater missions. However, learning perception and decision models individually might result in weak robustness of overall control system as the mismatched state extraction and control decision making are asynchronous. As a remedy, we will introduce in this paper an end-to-end monocular autonomous reinforcement control (MARC) framework for autonomous control of VUVs, which is performed in two cascaded procedures, i.e., 1) perception, where a geometric network (GeoNet) is designed based on a convolutional encoder-decoder network to generate depth maps from input environmental videos; 2) decision, where with depth maps as input, a reinforcement control network (CtrlNet) integrates a convolutional neural network into a deep deterministic policy gradient network and outputs action decisions, which are refined by reinforcement learning algorithm for obstacle-avoiding based autonomous control. Numerical and experimental results demonstrate that the proposed MARC exhibits high-quality depth prediction and is capable of conducting obstacle-avoiding navigation and autonomous control of VUVs with high accuracy and strong robustness.
- 24 May 2022
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Transactions on Vehicular Technology
- Vol. 71 (8) , 8237-8250
Autonomous Reinforcement Control of Visual Underwater Vehicles: Real-Time Experiments Using Computer Vision
Swift decision-making based on visual environment perception is crucial for autonomous control of visual underwater vehicles (VUVs) during underwater missions. However, learning perception and decision models individually might result in weak robustness of overall control system as the mismatched state extraction and control decision making are asynchronous. As a remedy, we will introduce in this paper an end-to-end monocular autonomous reinforcement control (MARC) framework for autonomous control of VUVs, which is performed in two cascaded procedures, i.e., 1) perception, where a geometric network (GeoNet) is designed based on a convolutional encoder-decoder network to generate depth maps from input environmental videos; 2) decision, where with depth maps as input, a reinforcement control network (CtrlNet) integrates a convolutional neural network into a deep deterministic policy gradient network and outputs action decisions, which are refined by reinforcement learning algorithm for obstacle-avoiding based autonomous control. Numerical and experimental results demonstrate that the proposed MARC exhibits high-quality depth prediction and is capable of conducting obstacle-avoiding navigation and autonomous control of VUVs with high accuracy and strong robustness.
- 23 November 2021
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE/ASME Transactions on Mechatronics
- Vol. 27 (5) , 2786-2795
An Active DC-Link Stabilization Strategy for Permanent Magnet Synchronous Motor Drive System Based on Inverter Voltage Compensation
Dc-link oscillation in electromechanical equipment drive system with film capacitor is a common problem when system impedance mismatch. For this problem, an inverter voltage compensation (IVC) strategy is proposed in this article to suppresses the oscillation of dc-link voltage. This method improves the motor input impedance by designing a damping compensation link so that the input and output impedances of the system can meet the stability conditions. It injects the damping signal into the inverter reference voltage of PWM modulation, which avoids the influence of current loop bandwidth and does not directly affect the motor speed or motor current. Moreover, the compensation impedance of this method is related to the motor voltage, which can compensate system damping dynamically according to the motor operating conditions, and the effect of the compensation link on the motor dynamic performance is reduced. Furthermore, the selection of compensation coefficients in this method is more intuitive and it simplifies the system design. Finally, experimental results demonstrate the effectiveness of the proposed method after building an experimental platform.
- 23 November 2021
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE/ASME Transactions on Mechatronics
- Vol. 27 (5) , 2786-2795
An Active DC-Link Stabilization Strategy for Permanent Magnet Synchronous Motor Drive System Based on Inverter Voltage Compensation
Dc-link oscillation in electromechanical equipment drive system with film capacitor is a common problem when system impedance mismatch. For this problem, an inverter voltage compensation (IVC) strategy is proposed in this article to suppresses the oscillation of dc-link voltage. This method improves the motor input impedance by designing a damping compensation link so that the input and output impedances of the system can meet the stability conditions. It injects the damping signal into the inverter reference voltage of PWM modulation, which avoids the influence of current loop bandwidth and does not directly affect the motor speed or motor current. Moreover, the compensation impedance of this method is related to the motor voltage, which can compensate system damping dynamically according to the motor operating conditions, and the effect of the compensation link on the motor dynamic performance is reduced. Furthermore, the selection of compensation coefficients in this method is more intuitive and it simplifies the system design. Finally, experimental results demonstrate the effectiveness of the proposed method after building an experimental platform.
- Open Access
- 26 January 2021
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 9, 18287-18297
Phase Shaping Method for Negative Input Admittance of Buck Converter Based on Sliding Mode Disturbance Observer
The power electronic converter that supplies power to the unmanned ship’s servers can cause self-excited oscillation of the LC input filter under heavy load. To suppress voltage oscillation, an input impedance shaping method based on a sliding mode disturbance observer is presented. The self-excited oscillation mechanism of the LC filter caused by the buck converter’s positive feedback is analyzed by using the small-signal model. The sliding mode disturbance observer calculates the input voltage according to the output voltage, inductance current and duty cycle of the buck converter. The feedforward compensation of the observed voltage reshapes the input admittance of the buck converter, making the system not meet the conditions of self-excited oscillation and suppressing the voltage oscillation of the LC filter. The phase shaping of input admittance is compared with the traditional amplitude shaping by the Bode diagram and Nyquist diagram. The experimental results show that the phase shaping method could better suppress the LC filter’s self-excited oscillation and the sliding mode disturbance observer reduced the hardware cost.
- 26 January 2021
- Published by Institute of Electrical and Electronics Engineers (IEEE) in IEEE Access
- Vol. 9, 18287-18297
Open AccessPhase Shaping Method for Negative Input Admittance of Buck Converter Based on Sliding Mode Disturbance Observer
The power electronic converter that supplies power to the unmanned ship’s servers can cause self-excited oscillation of the LC input filter under heavy load. To suppress voltage oscillation, an input impedance shaping method based on a sliding mode disturbance observer is presented. The self-excited oscillation mechanism of the LC filter caused by the buck converter’s positive feedback is analyzed by using the small-signal model. The sliding mode disturbance observer calculates the input voltage according to the output voltage, inductance current and duty cycle of the buck converter. The feedforward compensation of the observed voltage reshapes the input admittance of the buck converter, making the system not meet the conditions of self-excited oscillation and suppressing the voltage oscillation of the LC filter. The phase shaping of input admittance is compared with the traditional amplitude shaping by the Bode diagram and Nyquist diagram. The experimental results show that the phase shaping method could better suppress the LC filter’s self-excited oscillation and the sliding mode disturbance observer reduced the hardware cost.