Research direction：

Intelligent Unmanned System, Optoelectronic Unmanned System

Education and Experience：

12/2019 – Present, Associate Researcher, Department of Automation, Tsinghua University

09/2015 – 12/2019, Assistant Researcher, Department of Automation, Tsinghua University

01/2011 – 09/2015, Postdoc. at Department of Automation, Tsinghua University

09/2005 – 12/2010, Ph.D., Electronic Engineering, Tsinghua University

09/2003 – 07/2005, Master, Electronic Engineering, Harbin Institute of Technology

09/1999 – 07/2003, Bachelor, Electronic Engineering, Harbin Institute of Technology

Representative Publications in past 3 years (# co-first author, *corresponding author)：

1、Fan Zhen, Li Xiu,Li Yipeng*. Multi-Agent Deep Reinforcement Learning for Online 3D Human Poses Estimation. Remote Sensing, 2021, 13(19): 3995.

2、Li Siqi, Feng Yutong,Li Yipeng, Jiang Yu, Zou Changqing, Gao Yue. Event Stream Super-Resolution via Spatiotemporal Constraint Learning. Proceedings of the IEEE International Conference on Computer Vision, 2021: 4460-4469.

3、Zhou Tiankuang, Lin Xing, Wu Jiamin, Chen Yitong, Xie Hao,Li Yipeng,Fan Jintao, Wu Huaqiang, Fang Lu, Dai Qionghai. Large-scale neuromorphic optoelectronic computing with a reconfigurable diffractive processing unit, Nature Photonics, 2021, 15(5): 367-373.

4、Zhang Baosheng, Guo Yuchen,Li Yipeng, He Yuwei, Wang Haoqian, Dai Qionghai. Memory Recall: A Simple Neural Network Training Framework Against Catastrophic Forgetting, IEEE Transactions on Neural Networks and Learning Systems, 2022, 33(5): 2010-2022.

5、Yue Gao, Siqi Li,Yipeng Li, Yandong Guo, Qionghai Dai. SuperFast: 200 Video Frame Interpolation via Event Camera. IEEE Transactions on Pattern Analysis and Machine Intelligence. 2023, 45(6): 7764-7780.

Research Focus：

Intelligent drones will play an important role in low-altitude economic construction and national security. Traditional unmanned systems adopt a discrete architecture of "information collection + perception calculation + underlying control" to separate the collection of environmental information from perception calculation. Multiple sensors are required to collect data, synchronize and fuse together, resulting in slow perception speed and poor functionality. The bottlenecks of high consumption and heavy weight severely limit the intelligent development of unmanned systems. Compared with traditional silicon-based computing, optical computing has extremely high switching speed and transmission bandwidth, and has the advantages of high speed and high complexity. The rapid development of photoelectric computing theory provides the possibility for integrated information collection and calculation. In the future, we will focus on building an integrated theoretical architecture and method of photoelectric sensing and control, using "optical diffraction front-end + ultra-high-speed decoding decision-making" to replace the traditional discrete architecture. By establishing an end-to-end output system spanning from ambient light data to control parameters for unmanned systems, we aim to transcend the performance limitations imposed by existing architectures. This innovation promises to significantly enhance the autonomous flight capabilities of unmanned aerial systems, while also furnishing groundbreaking theoretical and technical foundations for the progressive advancement of unmanned systems as a whole.

Honors & Awards：

2023 Second Prize of Scientific and Technological Progress Award of China Electronics Society

2019 First Prize of Technical Invention Award of China Electronics Society

2015 First Prize of Scientific and Technological Progress Award of China Electronics Society

2013 World Champion in the 23^rdInternational Aerial Robot Competition

2012 Ranked fourth in the 22^ndInternational Aerial Robot Competition