博士（后），副教授，计算机学院副院长、博士生导师 

邮箱：xuewei@ahut.edu.cn

MILA（MachIne Learning and Applications）研究小组的研究方向为机器学习理论与算法，以及探索其在计算机视觉、机器视觉、医学图像处理、遥感图像处理等领域的交叉应用。

2025年拟招收3~4名硕士生。欢迎诚实守信、勤奋好学，态度端正，具有一定数学基础和编程能力的同学与我联系。邮件需包含以下内容：带近期生活照的个人简历、本科成绩单，以及能够体现个人综合能力的附件材料。

☆2024级硕士生☆

1. 马梦成（专硕）

2. 陈翔（专硕）

3. 何金峰（学硕）

☆2023级硕士生☆

1. 张雨晴（学硕）

One paper has been submitted to the Intelligent Data Analysis.

One paper has been submitted to the Statistics, Optimization and Information Computing.

2. 陈创慧（学硕）

One paper has been submitted to 电子学报.

Wei Xue, Chuanghui Chen, Xuan Qi, Jian Qin, Zhen Tang, Yongsheng He. M2ANet: Multi-branch and multi-scale attention network for medical image segmentation. Chinese Physics B, 2025, 34(8): 080703.

国家发明专利：基于多分支多尺度注意力机制的医学图像分割方法(申请号：202510499976.6)

3. 朱舒婷（专硕）

Wei Xue, Shuting Zhu, Yazhou Yao, Xiaozheng, Ping Zhong. Fusion of Infrared and Visible Images Based on Iterative Dual-branch Attention and Modality Discrepancy Guidance. Accepted by the IEEE Transactions on Multimedia.

国家发明专利：基于可迭代双分支注意力和模态差异引导的图像融合方法(申请号：202510556142.4)

4. 邵天诚（专硕）

One paper has been submitted to the IEEE Journal on Miniaturization for Air and Space Systems.

Wei Xue, Tiancheng Shao, Mingyang Du, Xiao Zheng, Ping Zhong. Partial Attention Feature Aggregation Network for Lightweight Remote Sensing Image Super-Resolution. IEEE Geoscience and Remote Sensing Letters, vol. 22, pp. 1-5, 2025, Art no. 6011905, doi: 10.1109/LGRS.2025.3601595.

国家发明专利：一种基于轻量级部分注意力卷积的遥感图像超分辨率方法(申请号：202510710197.6)

5. 曹贵猛（专硕）（co-advised by Prof. Chen Yebin）

Yebin Chen, Guimeng Cao, Wei Xue*. Infrared Dim Small Target Detection via Fractional-order TV and Nonnegative L1-norm. Engineering Letters, 2025, 33(8): 2925-2938.

国家发明专利：基于 FTV 与非负 L1 范数约束的红外弱小目标检测方法及系统(申请号：202510861650.3)

6. 桂晶晶（专硕）（co-advised by Prof. Chen Yebin）

One paper has been submitted to the Signal, Image and Video Processing.

Yebin Chen, Jingjing Gui, Wei Xue*. Overlapping Group Sparsity TV with Lp-Norm Constraint for Infrared Dim Small Target Detection. IAENG International Journal of Computer Science, 2025, 52(8): 2578-2592.

国家发明专利：基于 OGSTV 和 Lp 范数约束的红外弱小目标检测方法及系统(申请号：202510861653.7)

☆2022级硕士生☆

1. 王永皓（学硕）

Wei Xue, Yonghao Wang, Shaoquan Zhang, Xiao Zheng, Ping Zhong. On Adversarial Robust Generalization of DNNs for Remote Sensing Image Classiffcation. IEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, 2025, 18: 11370-11385.

Wei Xue, Yonghao Wang, Yuchi Wang, Yue Wang, Mingyang Du, Xiao Zheng. Towards a Better Trade-off between Accuracy and Robustness for Image Classification via Adversarial Feature Diversity. IEEE Journal on Miniaturization for Air and Space Systems, 2024, 5(4): 254-264.

国家发明专利：基于OAT框架的CNN标准精度与鲁棒精度的平衡方法(申请号：202410352969.9)

P.s. 获"2025届校优秀毕业生"荣誉称号；获安徽省高等学校计算机教育研究会2025年度优秀硕士学位论文奖

2. 李   婕（学硕）

薛伟, 李婕, 杜明洋. 梯度对齐下的目标检测模型对抗鲁棒性提升. 国防科技大学学报, 已录用.

Li J, Wang Y, and Xue W. Boosting the Adversarial Robustness for Object Detection via Adaptive Perturbation Control. Accepted by the 15th International Symposium on Intelligence Computation and Applications. (Best Paper Award)

国家发明专利：基于扰动幅度自适应控制的目标检测网络鲁棒性提升方法(申请号：202410360687.3)

3. 李   柔（学硕）

李柔, 薛伟. 基于L1范数与梯度约束的无人机图像拼接方法. 安徽工业大学学报(自然科学版), 已录用.

Li R and Xue W. Robust UAV Image Stitching via Multiscale Feature Matching and Local Homography Matrix Estimation. International Conference Automatics and Informatics, 2024, 358-363.

国家发明专利：基于局部扭曲和接缝引导的无人机图像拼接方法(申请号：202411044069.4 )

P.s. 获"2025届校优秀毕业生"荣誉称号

4. 韦娜娜（专硕）

Wei N, Xue W, Gu X, and Qi X. Magnetic Resonance Image Restoration by Utilizing Fractional-Order Total Variation and Recursive Filtering. Statistics, Optimization & Information Computing, 2025, 13: 138-157.

软件著作权：基于方差稳定变换的磁共振图像Rician噪声去除软件V1.0(登记号：2024SR0306067)

☆2021级硕士生☆

1. 许   赟（学硕）

Xu Y, Xue W, Zhu Q, et al. Gradient-guided Mechanism for Long-Tailed Medical Image Classification. China Automation Congress, 2023, 4475-4480.

国家发明专利：一种损失重平衡的方法、装置和电子设备(申请号：202311139754.0)

学位论文题目：基于损失再平衡的医学长尾图像分类方法研究

2. 段李龙（学硕）

Xue W, Duan L, Hong X, and Zheng X. Adaptive weighting and nearest neighbor-based area control for imbalanced data classification. Applied Soft Computing, 2025, 177: 113171.

Duan L, Xue W, Huang J, et al. Joint Sample Position-Based Noise Filtering and Mean Shift Clustering for Imbalanced Classification Learning. Tsinghua Science and Technology, 2024, 29(1): 216-231.

Duan L, Xue W, Gu X, et al. HSNF: Hybrid sampling with two-step noise filtering for imbalanced data classification. Intelligent Data Analysis, 2023, 27: 1573-1593.

国家发明专利：一种带有两步噪声过滤的不平衡数据混合采样算法(申请号：202310003668.0)

学位论文题目：基于重采样的类不平衡数据分类方法研究

P.s. 获研究生国家奖学金；获"2024届校优秀毕业生"荣誉称号；赴安徽大学攻读博士学位

3. 葛宇梦（专硕）

Xue W, Ge Y, Gu X, et al. Boosting MR Image Impulse Noise Removal with Overlapping Group Sparse Fractional-order Total Variation and Minimax Concave Penalty. Image Analysis and Stereology, 2024, 43: 53-66.

Ge Y, Xue W, Xu Y, et al. Magnetic Resonance Image Denoising Based On Laplacian Prior Sparsity Constraint And Nonconvex Second-Order TV Penalty. Image Analysis and Stereology, 2023, 42: 119-132.

软件著作权：基于欧氏度量和高阶全变分的颅脑MRI图像复原软件V1.0(登记号：2023SR0585061)

学位论文题目：基于重叠组稀疏全变分的MR图像复原方法研究

☆2020级硕士生☆

1. 夏筱彦（学硕）

Xue W, Xia X, Wan P, et al. Adversarial attack on object detection via object feature-wise attention and perturbation extraction. Accepted by the Tsinghua Science and Technology, 2025, 30(3): 1174-1189.

Xia X, Xue W, Wan P. FCGSM: Fast Conjugate Gradient Sign Method for Adversarial Attack on Image Classification. International Conference on Innovative Computing, 2023, 709-716.

国家发明专利：一种基于联合损失函数的图像对抗样本生成方法及系统(申请号：202210966497.7)

软件著作权：基于PyQT5的目标检测对抗攻击演示系统V1.0(登记号：2022SR0825115)

学位论文题目：深度神经网络中的图像识别对抗攻击方法研究

P.s. 赴国防科技大学攻读博士学位

2. 魏雪玉（学硕）

Wei X, Duan L and Xue W. Reward-weighted DHER Mechanism For Multi-goal Reinforcement Learning With Application To Robotic Manipulation Control. Journal of Applied Science and Engineering, 2023, 26(12): 1829-1841.

Wei X, Xue W, Zhao W, et al. Joint Action Representation and Prioritized Experience Replay for Reinforcement Learning in Large Discrete Action Spaces. International Conference on Machine Learning and Soft Computing, 2023, 96-101.

国家发明专利：用于大型离散动作空间的强化学习优化方法、方法及装置(专利号：ZL202210655295.0；授权公告号：CN114925778B)

学位论文题目：连续状态下基于经验回放的强化学习方法研究

P.s. 获"2023届校优秀毕业生"荣誉称号

3. 郭事成（学硕）

Guo S, Xue W, Zhao W, et al. Efficient Exploitation for Off-policy Reinforcement Learning via a Novel Multi-scale Tabular Buffer. Asia Conference on Information Engineering, 2023, 1-5.

Guo S, Wei X, Xu Y, et al. A Survey of Linear Value Function Approximation in Reinforcement Learning. International Symposium on Intelligence Computation and Applications, 2021, 266–280.

学位论文题目：基于样本池的深度强化学习样本效率提升研究

4. 张海宏（学硕）（co-advised by Prof. Tao Tao）

Xue W, Zhang H, Wei X, et al. Optimizing Exploration-Exploitation Trade-off in Continuous Action Spaces via Q-ensemble. Pacific Rim International Conference on Artificial Intelligence, 2022, 148-160.

国家发明专利：一种集成多价值网络的强化学习算法、装置及电子设备(申请号：202210653373.3)

学位论文题目：连续空间下值函数高估偏差修正方法研究

☆2019级硕士生☆

1. 王孝文（学硕）

王孝文, 李乔, 薛伟, 钟平. 联合γ-范数和TV-稀疏约束的红外弱小目标检测. 航空兵器, 2022, 29(2): 30-38.

Wang X, Xia X, Li Q, et al. Infrared dim and small target detection based on total variation and multiple noise constraints modeling. International Conference on Image Processing and Machine Vision, 2022, 1-5.

学位论文题目：基于图像数据结构的红外弱小目标检测方法研究

2. 陈学灿（专硕）

软件著作权：基于C/S架构的病患医学影像诊断数据管理平台V1.0(登记号：2022SR0210306)

软件著作权：基于B/S架构的输液泵监测系统V1.0(登记号：2022SR0582732)

学位论文题目：基于LoRa和Django框架的输液泵监测系统设计与实现

3. 万鹏程（学硕）（co-advised by Prof. Tao Tao）

Xue W, Wan P, Li Q, et al. An online conjugate gradient algorithm for large-scale data analysis in machine learning. AIMS Mathematics, 2021, 6(2): 1515-1537.

Wan P, Mu Y, Xue W, et al. Feedforward Neural Network Based on Nonmonotone Conjugate Gradient Method for Multiple Classification. International Joint Conference on Neural Networks, 2021.

学位论文题目：基于共轭梯度的机器学习分类方法研究

P.s. 获研究生国家奖学金；赴国防科技大学攻读博士学位

☆2016级硕士生☆

邵光梅（学硕）（co-advised by Prof. Zheng Xiao）

Shao G, Xue W, Yu G, et al. Improved SVRG for finite sum structure optimization with application to binary classification. Journal of Industrial and Management Optimization, 2020, 16(5): 2253-2266.

学位论文题目：基于深度卷积网络的板带钢表面缺陷类别检测.