过去的事就过去吧。
I am a computational mathematics researcher working on artificial intelligence model design, mathematical modeling, and experimental validation. My research experience spans neural network architectures, computer vision, spiking neural networks, model robustness, and continuous-time generative modeling.
I am especially interested in building AI methods that connect mathematical structure, model implementation, and real-world data-driven applications.
- Ph.D. in Computational Mathematics, Jilin University.
- M.S.-Ph.D. track in Fundamental Mathematics and Computational Mathematics, Jilin University.
- B.S. in Mathematics and Applied Mathematics, Jilin University.
- 🎵 Classical Music and Pink Floyd.
- 📚 Marxism and Maoism.
- 💻 Coding and Gaming.
- 🧘 Focusing and Emptying.
- 📘 Reading and Writing.
- 💼 Working and Relaxing.
- ❤️ Favourite programming language:
C#. - 👨💻 Programming languages:
Python,C/C++,C#,Rust,HTML,LaTeX. - 😀 AI and data tools:
PyTorch,Lightning,Scikit-learn,NumPy. - 🔧 Research engineering: model implementation, CUDA operator development, experiment scripting, debugging, result analysis, and technical writing.
- 🌟 Embodied Intelligence and World Models / 具身智能与世界模型
- 🌟 Vision-Language Models and Vision-Language-Action Models / 视觉语言模型与视觉语言动作模型
- 🌟 Neural Differential Equations and Flow Matching / 神经微分方程与流匹配
- 🌟 Spiking Neural Networks / 脉冲神经网络
- 🌟 Computer Vision / 计算机视觉
- Neural network architecture and representation learning: model structure design, feature representation analysis, model explanation, and experimental comparison.
- Temporal modeling, latency optimization, and robustness: adversarial attacks, dynamic training, temporal coding, and efficiency analysis for spiking neural networks.
- Computer vision and interdisciplinary AI: image classification, object detection, segmentation, medical image analysis, remote sensing road extraction, and change detection.
- Continuous-time modeling and generative dynamics: neural differential equations, Flow Matching, data-driven dynamical systems, time-series prediction, and scientific computing.
- Research workflow: literature reading, model reproduction, code implementation, training scripts, ablation studies, result analysis, paper writing, and technical documentation.
-
Xiao Du, Wanli Shi, Xiaohan Zhao, Yang Cao, Bin Gu, and Tieru Wu. “Raw Event-Based Adversarial Attacks for Spiking Neural Networks with Configurable Latencies.” Neural Networks, Volume 193, 2026, 108026, ISSN 0893-6080, https://doi.org/10.1016/j.neunet.2025.108026.
-
Xiao Du, Wanli Shi, Hanyuan Zheng, Bhaskar Mukhoty, Yang Cao, Bin Gu, and Tieru Wu. “Dynamic Training of Spiking Neural Networks with Loss-Based Stochastic Latency.” Computer Vision and Image Understanding, 2026, 104873, ISSN 1077-3142, https://doi.org/10.1016/j.cviu.2026.104873.
-
Jinjie Fang, Xiao Du, Tianxing Man, Chengxun Jin, Haozhen Zhang, Yi Chang, and Bin Gu. “Achieve Latency-Efficient Temporal-Coding Spiking LLMs via Discretization-Aware Conversion.” Submitted to NeurIPS 2026. https://openreview.net/forum?id=zrGcuTNwu1.
- Xiao Du, Ziyou Guo, Zihao Li, Yang Cao, Xing Chen, and Tieru Wu. “VexNet: Vector-Composed Feature-Oriented Neural Network.” Electronics. 2025; 14(9):1897. https://doi.org/10.3390/electronics14091897.
- Shuaichen Liu, Xiao Du, and Tieru Wu. “Stacked Multi-Head Cross Attention for Image Recognition.” In Proceedings of the International Conference on Image Processing, Machine Learning and Pattern Recognition (IPMLP '24). Association for Computing Machinery, New York, NY, USA, 376–381. https://doi.org/10.1145/3700906.3700965.
- Ruyue Feng, Ziyou Guo, Xiao Du, and Tieru Wu. “SAM2-RoadNet: Topology-Aware Multi-Scale Road Extraction from High-Resolution Remote Sensing Images.” Remote Sensing. 2026; 18(6):913. https://doi.org/10.3390/rs18060913.
-
Zihao Li, Xiao Du, Ao Xu, Tieru Wu, and Yang Cao. “Explaining Tree Ensembles through Single Decision Trees.” Information Fusion, Volume 123, 2025, 103244, ISSN 1566-2535, https://doi.org/10.1016/j.inffus.2025.103244.
-
Zihao Li, Xiao Du, Tieru Wu, and Yang Cao. “Explaining Random Forests as Single Decision Trees through Distance Functional Optimization.” 2024 International Joint Conference on Neural Networks (IJCNN), Yokohama, Japan, 2024, pp. 1-8, doi: 10.1109/IJCNN60899.2024.10650261.
- Chen Zhang, Ziyou Guo, Weixin Bu, Xiao Du, Runsheng Yu, Yik-Chung Wu, and Ngai Wong. “Nonparametric Teaching of Sequential Properties in Recurrent Neural Learners.” Submitted to NeurIPS 2026. https://openreview.net/forum?id=rQIzp565NP.
- Chen Zhang, Xiao Du, Ziyou Guo, Weixin Bu, Tieru Wu, Yik-Chung Wu, and Ngai Wong. “Accelerating Graph Property Learning via Evolutionary Sample Selection.” Submitted to NeurIPS 2026. https://openreview.net/forum?id=IzSGuwmz7I.
- ⭐ Convolutional Neural Works
- ⭐ Generative Adversarial Networks
- ⭐ Image Classification, Object Detection, Instance/Semantic Segmentation.
- ⭐ Capsule Networks.
- 🏫 School of Mathematics，School of Artificial Intelligence，Jilin University.
📧 onbigion@gmail.com; onbigion@live.com.
🐦 G.t 霁天 (@GtChoc) / X.
📘 霁天(@onbigion13)