Yu Sun
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
Research Subjects
Yu Sun has a total of 33 co-authors. Here are the top co-authors by number of shared publications:
- Hongda Yue
School of Computer and Electronics and Information, Guangxi University, Nanning, China
3 shared publications - Bei Hua
School of Computer, Electronics and Information, Guangxi University, Nanning 530004, China
2 shared publications - Jian Feng
School of Computer and Electronics and Information, Guangxi University, Nanning, People’s Republic of China
2 shared publications - Yingying Yang
School of Computer and Electronics and Information, Guangxi University, Nanning 530004, China
2 shared publications - Bei Hua
School of Computer and Electronics Information, Guangxi University, Nanning 530004, China
1 shared publication - Bozhao Chen
School of Computer and Electronics and Information, Guangxi University, Nanning 530004, China
1 shared publication
Latest Publications
- Open Access
- 10 July 2026
- Published by Association for Computing Machinery (ACM)
- p. 419-427
An Adaptive Offspring Generation Wrapper for Constrained Multimodal Multiobjective Evolutionary Algorithms
Constrained multimodal multiobjective optimization problems (CM-MOPs) require balancing multiple conflicting objectives under complex constraints while achieving comprehensive coverage of multiple disjoint constrained Pareto sets (CPSs) in the decision space. Since offspring generation is crucial for exploring and maintaining these CPSs, this paper proposes a new constrained multimodal multiobjective evolutionary algorithm, termed CMMEA-AOGW. CMMEA-AOGW incorporates an adaptive offspring generation wrapper (AOGW) that dynamically switches between the baseline offspring generator and two complementary sparse interpolation operators: skewness-guided sparse interpolation (SGSI) and sparse interpolation oriented to objective space (SIOS). SGSI identifies sparse and boundary regions in the decision space by analyzing the skewness of neighborhood distance distributions and generates offspring via boundary extrapolation and directional interpolation, thereby improving CPS coverage and preserving multimodal diversity. In addition, SIOS detects low-density gaps along the constrained Pareto front (CPF) in the objective space and performs interpolation-based compensation to enhance CPF completeness and uniformity. Experimental results demonstrate that CMMEA-AOGW outperforms state-of-the-art CMMEAs and ranks first in the Friedman average rankings across four performance metrics. Moreover, embedding studies on multiple representative CMMEAs validate the effectiveness and robustness of AOGW, showing consistent performance improvements across different baselines.
- 10 July 2026
- Published by Association for Computing Machinery (ACM)
- p. 419-427
Open AccessAn Adaptive Offspring Generation Wrapper for Constrained Multimodal Multiobjective Evolutionary Algorithms
Constrained multimodal multiobjective optimization problems (CM-MOPs) require balancing multiple conflicting objectives under complex constraints while achieving comprehensive coverage of multiple disjoint constrained Pareto sets (CPSs) in the decision space. Since offspring generation is crucial for exploring and maintaining these CPSs, this paper proposes a new constrained multimodal multiobjective evolutionary algorithm, termed CMMEA-AOGW. CMMEA-AOGW incorporates an adaptive offspring generation wrapper (AOGW) that dynamically switches between the baseline offspring generator and two complementary sparse interpolation operators: skewness-guided sparse interpolation (SGSI) and sparse interpolation oriented to objective space (SIOS). SGSI identifies sparse and boundary regions in the decision space by analyzing the skewness of neighborhood distance distributions and generates offspring via boundary extrapolation and directional interpolation, thereby improving CPS coverage and preserving multimodal diversity. In addition, SIOS detects low-density gaps along the constrained Pareto front (CPF) in the objective space and performs interpolation-based compensation to enhance CPF completeness and uniformity. Experimental results demonstrate that CMMEA-AOGW outperforms state-of-the-art CMMEAs and ranks first in the Friedman average rankings across four performance metrics. Moreover, embedding studies on multiple representative CMMEAs validate the effectiveness and robustness of AOGW, showing consistent performance improvements across different baselines.
- Open Access
- 12 June 2026
- Published by MDPI AG in Electronics
- Vol. 15 (12) , 2599
ESKF-g2o-SLAM: A Stereo Visual–Inertial SLAM with ORB Features and ESKF-Based VIO
With the development of the low-altitude economy, low-altitude intelligent agents such as delivery robots, courier drones, and outdoor cleaning robots are gradually moving towards widespread application. One of the core challenges faced by such systems is localization and mapping in complex scenarios characterized by satellite signal denial and unknown environmental prior information. To address this requirement, this paper proposes ESKF-g2o-SLAM, a stereo visual-inertial SLAM system that integrates an ESKF (Error-State Kalman Filter)-based visual-inertial odometry front-end with an ORB-feature-based g2o graph optimization back-end in a cascaded, loosely coupled manner. The proposed method was evaluated on 11 sequences of the EuRoC dataset and compared with state-of-the-art approaches including ORB-SLAM2 (stereo), MSCKF-VIO, OKVIS, and VINS-Fusion (stereo). Ablation studies show marginal improvements on selected sequences and suggest potential robustness advantages under more challenging visual conditions. Experimental results show that our method achieves competitive accuracy in terms of both Absolute Trajectory Error (ATE) and Relative Pose Error (RPE), exhibiting good robustness and stability.
- 12 June 2026
- Published by MDPI AG in Electronics
- Vol. 15 (12) , 2599
Open AccessESKF-g2o-SLAM: A Stereo Visual–Inertial SLAM with ORB Features and ESKF-Based VIO
With the development of the low-altitude economy, low-altitude intelligent agents such as delivery robots, courier drones, and outdoor cleaning robots are gradually moving towards widespread application. One of the core challenges faced by such systems is localization and mapping in complex scenarios characterized by satellite signal denial and unknown environmental prior information. To address this requirement, this paper proposes ESKF-g2o-SLAM, a stereo visual-inertial SLAM system that integrates an ESKF (Error-State Kalman Filter)-based visual-inertial odometry front-end with an ORB-feature-based g2o graph optimization back-end in a cascaded, loosely coupled manner. The proposed method was evaluated on 11 sequences of the EuRoC dataset and compared with state-of-the-art approaches including ORB-SLAM2 (stereo), MSCKF-VIO, OKVIS, and VINS-Fusion (stereo). Ablation studies show marginal improvements on selected sequences and suggest potential robustness advantages under more challenging visual conditions. Experimental results show that our method achieves competitive accuracy in terms of both Absolute Trajectory Error (ATE) and Relative Pose Error (RPE), exhibiting good robustness and stability.
- Open Access
- 15 May 2026
- Published by Association for Computing Machinery (ACM)
- p. 305-312
MOEA/D-EPS: Evolutionary-Potential Scheduling for Decomposition-Based Multiobjective Optimization
- 15 May 2026
- Published by Association for Computing Machinery (ACM)
- p. 305-312
Open AccessMOEA/D-EPS: Evolutionary-Potential Scheduling for Decomposition-Based Multiobjective Optimization
- Open Access
- 15 May 2026
- Published by Association for Computing Machinery (ACM)
- p. 42-48
DA-CTransNet: A Transformer-CNN Hybrid Framework for Dermoscopy ImageSegmentation via Domain Adaptation and Cascaded Boundary Refinement
Accurate skin lesion segmentation is a critical foundation for computer-aided dermoscopic diagnosis. Existing CNN-Transformer hybrid methods still fall short in achieving deep fusion of global semantics and local details, as well as in task-specific boundary guidance mechanisms, especially given the core challenges in dermoscopy images: highly blurred lesion boundaries, complex and variable internal textures, and ubiquitous interference from hair occlusion and illumination artifacts. To address these issues, this paper proposes a novel segmentation network named DA-CTransNet, which deeply integrates three core modules: domain-knowledge-aware adaptive attention, multi-scale collaborative affine calibration, and cascaded boundary refinement. Specifically: First, we construct a domain-knowledge adaptive module that encodes clinically validated key visual diagnostic patterns (e.g., textures and vascular structures of specific lesions) into learnable feature prompt vectors via a lightweight sub-network, and adaptively injects them into the self-attention computation of the Transformer to achieve targeted enhancement of semantic features. Second, we propose a multi-scale collaborative affine calibration mechanism that, based on cross-spatial and cross-channel attention interaction, explicitly estimates and compensates for statistical distribution and spatial geometric discrepancies among features from different levels, constructing an intrinsically consistent and aligned feature pyramid. Finally, we design a dual-pathway cascaded boundary refinement decoder: the global pathway performs coarse segmentation, while the boundary pathway introduces a spectral-aware uncertainty mapping module to locate blurred boundary regions, combined with a high-order edge-preserving loss for end-to-end learning to achieve sub-pixel boundary reconstruction. Extensive experiments on the ISIC 2017 and ISIC 2018 datasets demonstrate that, compared to the baseline model, DA-CTransNet improves the Dice coefficient by approximately 2.3%, with significant improvements in Boundary F1 score and HD95 metrics; compared to existing mainstream methods, the Hausdorff distance is reduced by approximately 15%, and the model exhibits stronger robustness on images containing artifacts. This study provides a new technical pathway for building high-performance clinical intelligent auxiliary systems.
- 15 May 2026
- Published by Association for Computing Machinery (ACM)
- p. 42-48
Open AccessDA-CTransNet: A Transformer-CNN Hybrid Framework for Dermoscopy ImageSegmentation via Domain Adaptation and Cascaded Boundary Refinement
Accurate skin lesion segmentation is a critical foundation for computer-aided dermoscopic diagnosis. Existing CNN-Transformer hybrid methods still fall short in achieving deep fusion of global semantics and local details, as well as in task-specific boundary guidance mechanisms, especially given the core challenges in dermoscopy images: highly blurred lesion boundaries, complex and variable internal textures, and ubiquitous interference from hair occlusion and illumination artifacts. To address these issues, this paper proposes a novel segmentation network named DA-CTransNet, which deeply integrates three core modules: domain-knowledge-aware adaptive attention, multi-scale collaborative affine calibration, and cascaded boundary refinement. Specifically: First, we construct a domain-knowledge adaptive module that encodes clinically validated key visual diagnostic patterns (e.g., textures and vascular structures of specific lesions) into learnable feature prompt vectors via a lightweight sub-network, and adaptively injects them into the self-attention computation of the Transformer to achieve targeted enhancement of semantic features. Second, we propose a multi-scale collaborative affine calibration mechanism that, based on cross-spatial and cross-channel attention interaction, explicitly estimates and compensates for statistical distribution and spatial geometric discrepancies among features from different levels, constructing an intrinsically consistent and aligned feature pyramid. Finally, we design a dual-pathway cascaded boundary refinement decoder: the global pathway performs coarse segmentation, while the boundary pathway introduces a spectral-aware uncertainty mapping module to locate blurred boundary regions, combined with a high-order edge-preserving loss for end-to-end learning to achieve sub-pixel boundary reconstruction. Extensive experiments on the ISIC 2017 and ISIC 2018 datasets demonstrate that, compared to the baseline model, DA-CTransNet improves the Dice coefficient by approximately 2.3%, with significant improvements in Boundary F1 score and HD95 metrics; compared to existing mainstream methods, the Hausdorff distance is reduced by approximately 15%, and the model exhibits stronger robustness on images containing artifacts. This study provides a new technical pathway for building high-performance clinical intelligent auxiliary systems.
- Open Access
- 28 April 2026
- Published by MDPI AG in Electronics
- Vol. 15 (9) , 1870
Multi-Directional Guided Dual-Mode Kriging-Assisted Competitive Particle Swarm Optimization
Surrogate-assisted evolutionary algorithms have become the mainstream approach for solving expensive constrained multi-objective optimization problems (ECMOPs). However, existing methods suffer from blind search issues, and their selection strategies fail to adapt to changes in evolutionary stages. To overcome these limitations, this paper proposes a Multi-directional Guided Dual-mode Kriging-assisted Competitive Particle Swarm Optimization (MGD-KCSO) algorithm. MGD-KCSO integrates three synergistic strategies: a multi-directional guided solution strategy that constructs four complementary search paths based on non-dominated solutions to effectively enhance convergence and diversity; a dual-population data selection strategy that separates unconstrained and constrained populations to perform objective-oriented and constraint-oriented optimization, respectively; and an adaptive infill sampling strategy that dynamically switches sampling modes by monitoring the change rate of the objective function of the ideal point. If this rate exceeds a predefined threshold, the algorithm executes unconstrained sampling to accelerate convergence; otherwise, it switches to constrained sampling to prioritize the exploration of feasible boundaries. To verify the effectiveness of MGD-KCSO, comprehensive experiments were conducted on 33 benchmark problems and two real-world engineering design problems (pressure vessel and disc brake design). MGD-KCSO was compared against eight classic algorithms and three state-of-the-art methods published in the past two years. Experimental results evaluated by inverted generational distance (IGD) and hypervolume (HV) metrics demonstrate that MGD-KCSO outperforms the comparative algorithms on most test instances, achieving superior performance in terms of convergence, diversity, and practical applicability.
- 28 April 2026
- Published by MDPI AG in Electronics
- Vol. 15 (9) , 1870
Open AccessMulti-Directional Guided Dual-Mode Kriging-Assisted Competitive Particle Swarm Optimization
Surrogate-assisted evolutionary algorithms have become the mainstream approach for solving expensive constrained multi-objective optimization problems (ECMOPs). However, existing methods suffer from blind search issues, and their selection strategies fail to adapt to changes in evolutionary stages. To overcome these limitations, this paper proposes a Multi-directional Guided Dual-mode Kriging-assisted Competitive Particle Swarm Optimization (MGD-KCSO) algorithm. MGD-KCSO integrates three synergistic strategies: a multi-directional guided solution strategy that constructs four complementary search paths based on non-dominated solutions to effectively enhance convergence and diversity; a dual-population data selection strategy that separates unconstrained and constrained populations to perform objective-oriented and constraint-oriented optimization, respectively; and an adaptive infill sampling strategy that dynamically switches sampling modes by monitoring the change rate of the objective function of the ideal point. If this rate exceeds a predefined threshold, the algorithm executes unconstrained sampling to accelerate convergence; otherwise, it switches to constrained sampling to prioritize the exploration of feasible boundaries. To verify the effectiveness of MGD-KCSO, comprehensive experiments were conducted on 33 benchmark problems and two real-world engineering design problems (pressure vessel and disc brake design). MGD-KCSO was compared against eight classic algorithms and three state-of-the-art methods published in the past two years. Experimental results evaluated by inverted generational distance (IGD) and hypervolume (HV) metrics demonstrate that MGD-KCSO outperforms the comparative algorithms on most test instances, achieving superior performance in terms of convergence, diversity, and practical applicability.
- Open Access
- 30 January 2026
- Published by MDPI AG in Electronics
- Vol. 15 (3) , 616
Large-Scale Sparse Multimodal Multiobjective Optimization via Multi-Stage Search and RL-Assisted Environmental Selection
Multimodal multiobjective optimization problems (MMOPs) are widely encountered in real-world applications. While numerous evolutionary algorithms have been developed to locate equivalent Pareto-optimal solutions, existing Multimodal Multiobjective Evolutionary Algorithms (MMOEAs) often struggle to handle large-scale decision variables and sparse Pareto sets due to the curse of dimensionality and unknown sparsity. To address these challenges, this paper proposes a novel approach named MASR-MMEA, which stands for Large-scale Sparse Multimodal Multiobjective Optimization via Multi-stage Search and Reinforcement Learning (RL)-assisted Environmental Selection. Specifically, to enhance search efficiency, a multi-stage framework is established incorporating three key innovations. First, a dual-strategy genetic operator based on improved hybrid encoding is designed, employing sparse-sensing dynamic redistribution for binary vectors and a sparse fuzzy decision framework for real vectors. Second, an affinity-based elite strategy utilizing Mahalanobis distance is introduced to pair real vectors with compatible binary vectors, increasing the probability of generating superior offspring. Finally, an adaptive sparse environmental selection strategy assisted by Multilayer Perceptron (MLP) reinforcement learning is developed. By utilizing the MLP-generated Guiding Vector (GDV) to direct the evolutionary search toward efficient regions and employing an iteration-based adaptive mechanism to regulate genetic operators, this strategy accelerates convergence. Furthermore, it dynamically quantifies population-level sparsity and adjusts selection pressure through a modified crowding distance mechanism to filter structural redundancy, thereby effectively balancing convergence and multimodal diversity. Comparative studies against six state-of-the-art methods demonstrate that MASR-MMEA significantly outperforms existing approaches in terms of both solution quality and convergence speed on large-scale sparse MMOPs.
- 30 January 2026
- Published by MDPI AG in Electronics
- Vol. 15 (3) , 616
Open AccessLarge-Scale Sparse Multimodal Multiobjective Optimization via Multi-Stage Search and RL-Assisted Environmental Selection
Multimodal multiobjective optimization problems (MMOPs) are widely encountered in real-world applications. While numerous evolutionary algorithms have been developed to locate equivalent Pareto-optimal solutions, existing Multimodal Multiobjective Evolutionary Algorithms (MMOEAs) often struggle to handle large-scale decision variables and sparse Pareto sets due to the curse of dimensionality and unknown sparsity. To address these challenges, this paper proposes a novel approach named MASR-MMEA, which stands for Large-scale Sparse Multimodal Multiobjective Optimization via Multi-stage Search and Reinforcement Learning (RL)-assisted Environmental Selection. Specifically, to enhance search efficiency, a multi-stage framework is established incorporating three key innovations. First, a dual-strategy genetic operator based on improved hybrid encoding is designed, employing sparse-sensing dynamic redistribution for binary vectors and a sparse fuzzy decision framework for real vectors. Second, an affinity-based elite strategy utilizing Mahalanobis distance is introduced to pair real vectors with compatible binary vectors, increasing the probability of generating superior offspring. Finally, an adaptive sparse environmental selection strategy assisted by Multilayer Perceptron (MLP) reinforcement learning is developed. By utilizing the MLP-generated Guiding Vector (GDV) to direct the evolutionary search toward efficient regions and employing an iteration-based adaptive mechanism to regulate genetic operators, this strategy accelerates convergence. Furthermore, it dynamically quantifies population-level sparsity and adjusts selection pressure through a modified crowding distance mechanism to filter structural redundancy, thereby effectively balancing convergence and multimodal diversity. Comparative studies against six state-of-the-art methods demonstrate that MASR-MMEA significantly outperforms existing approaches in terms of both solution quality and convergence speed on large-scale sparse MMOPs.
- 13 August 2025
- Published by Elsevier in Sensors and Actuators B: Chemical
- Vol. 445
CEGNN: Contrastive edge labeling graph neural network for sensor drift adaptation in electronic noses
- 13 August 2025
- Published by Elsevier in Sensors and Actuators B: Chemical
- Vol. 445
CEGNN: Contrastive edge labeling graph neural network for sensor drift adaptation in electronic noses
- Open Access
- 24 June 2025
- Published by MDPI AG in Applied Sciences
- Vol. 15 (13) , 7095
Precise Feature Removal Method Based on Semantic and Geometric Dual Masks in Dynamic SLAM
In visual Simultaneous Localization and Mapping (SLAM) systems, dynamic elements in the environment pose significant challenges that complicate reliable feature matching and accurate pose estimation. To address the issue of unstable feature points within dynamic regions, this study proposes a robust dual-mask filtering strategy that synergistically integrates semantic segmentation information with geometric outlier detection techniques. The proposed method first identifies outlier feature points through rigorous geometric consistency checks, then employs morphological dilation to expand the initially detected dynamic regions. Subsequently, the expanded mask is intersected with instance-level semantic segmentation results to precisely delineate dynamic areas, effectively constraining the search space for feature matching and reducing interference caused by dynamic objects. A key innovation of this approach is the incorporation of a Perspective-n-Point (PnP)-based optimization module. This module dynamically updates the outlier set on a per-frame basis, enabling continuous monitoring and selective removal of dynamic features. Extensive experiments conducted on benchmark datasets demonstrate that the proposed method achieves average accuracy improvements of 3.43% and 11.42% on the KITTI dataset and 24% and 8.27% on the TUM dataset. Compared to traditional methods, this dual-mask collaborative filtering strategy improves the accuracy of dynamic feature removal and enhances the reliability of dynamic object detection, validating its robustness and applicability in complex dynamic environments.
- 24 June 2025
- Published by MDPI AG in Applied Sciences
- Vol. 15 (13) , 7095
Open AccessPrecise Feature Removal Method Based on Semantic and Geometric Dual Masks in Dynamic SLAM
In visual Simultaneous Localization and Mapping (SLAM) systems, dynamic elements in the environment pose significant challenges that complicate reliable feature matching and accurate pose estimation. To address the issue of unstable feature points within dynamic regions, this study proposes a robust dual-mask filtering strategy that synergistically integrates semantic segmentation information with geometric outlier detection techniques. The proposed method first identifies outlier feature points through rigorous geometric consistency checks, then employs morphological dilation to expand the initially detected dynamic regions. Subsequently, the expanded mask is intersected with instance-level semantic segmentation results to precisely delineate dynamic areas, effectively constraining the search space for feature matching and reducing interference caused by dynamic objects. A key innovation of this approach is the incorporation of a Perspective-n-Point (PnP)-based optimization module. This module dynamically updates the outlier set on a per-frame basis, enabling continuous monitoring and selective removal of dynamic features. Extensive experiments conducted on benchmark datasets demonstrate that the proposed method achieves average accuracy improvements of 3.43% and 11.42% on the KITTI dataset and 24% and 8.27% on the TUM dataset. Compared to traditional methods, this dual-mask collaborative filtering strategy improves the accuracy of dynamic feature removal and enhances the reliability of dynamic object detection, validating its robustness and applicability in complex dynamic environments.
- Open Access
- 27 April 2025
- Published by MDPI AG in Applied Sciences
- Vol. 15 (9) , 4847
A Preference Model-Based Surrogate-Assisted Constrained Multi-Objective Evolutionary Algorithm for Expensively Constrained Multi-Objective Problems
In the context of expensive constraint multi-objective problems, it is evident that the feasible domain shapes and sizes of different problems vary considerably. The difficulty in finding optimal solutions presents a significant challenge in ensuring the surrogate-assisted evolutionary algorithm’s feasibility, convergence, and diversity. To more effectively address the distinctive characteristics of the feasible domain and objective function across a range of problems, we have developed a Kriging-based surrogate-assisted evolutionary algorithm tailored to the current population’s preferences. The algorithm can optimize the population according to the current population’s requirements. Additionally, considering the varying degrees of accuracy observed in the surrogate models at different stages, this paper employs a dynamic approach to the number of surrogate model evaluations, contingent on the accuracy of the current surrogate model. Two types of Pareto frontier search are distinguished: unconstrained and constrained. Moreover, distinct fill sampling strategies are devised in accordance with the specific optimization requirements of the current population. After assessing the proposed solutions, the discrepancy between the actual fitness value and the surrogate model’s prediction is calculated.The discrepancy is used to modify the number of evaluations conducted on the surrogate model. In order to illustrate the algorithm’s efficacy, it is benchmarked against the current state-of-the-art algorithms on various test problems. The experimental results demonstrate that the proposed algorithm performs better than other advanced methods.
- 27 April 2025
- Published by MDPI AG in Applied Sciences
- Vol. 15 (9) , 4847
Open AccessA Preference Model-Based Surrogate-Assisted Constrained Multi-Objective Evolutionary Algorithm for Expensively Constrained Multi-Objective Problems
In the context of expensive constraint multi-objective problems, it is evident that the feasible domain shapes and sizes of different problems vary considerably. The difficulty in finding optimal solutions presents a significant challenge in ensuring the surrogate-assisted evolutionary algorithm’s feasibility, convergence, and diversity. To more effectively address the distinctive characteristics of the feasible domain and objective function across a range of problems, we have developed a Kriging-based surrogate-assisted evolutionary algorithm tailored to the current population’s preferences. The algorithm can optimize the population according to the current population’s requirements. Additionally, considering the varying degrees of accuracy observed in the surrogate models at different stages, this paper employs a dynamic approach to the number of surrogate model evaluations, contingent on the accuracy of the current surrogate model. Two types of Pareto frontier search are distinguished: unconstrained and constrained. Moreover, distinct fill sampling strategies are devised in accordance with the specific optimization requirements of the current population. After assessing the proposed solutions, the discrepancy between the actual fitness value and the surrogate model’s prediction is calculated.The discrepancy is used to modify the number of evaluations conducted on the surrogate model. In order to illustrate the algorithm’s efficacy, it is benchmarked against the current state-of-the-art algorithms on various test problems. The experimental results demonstrate that the proposed algorithm performs better than other advanced methods.