Research experience at Peking University

I have worked as a research assistant on autonomous driving techniques (e.g., semantic perception, interactive behavior learning) from 2018/09 to 2020/07, under the supervision of Prof. Huijing Zhao, and completed my Bachelor thesis there.

Here are the research projects completed during this time:

• Proposed the optimal composition and combination of data input from various sensor types for the semantic segmentation of 3D LiDAR data, crucial for semantic understanding in autonomous driving.
• Implemented Semi-Supervised Learning for Semantic segmentation of 3D LiDAR data in dynamic scenes, achieving comparable performance using only 50% labeled data.
• Constructed a simulator using CARLA to model interactions between autonomous vehicles and pedestrians, emulating crucial decision-making scenarios in autonomous driving, and employed (Inverse) Reinforcement Learning techniques to facilitate driving behavior learning.
• Bachelor Thesis: Autonomous Driving Control for Interaction with Pedestrains based on Imitation Learing