Xinyi Liu

Xinyi Liu

Technical University of Denmark

👋 I am a final year M.S. student at the Technical University of Denmark with the Full Tuition Fee Waiver (30 per year). Previously, I was an intern in computer vision at Teton and Volvo Cars, and I am a research assistant at UCLA. I received my bachelor’s degree from Tongji University with the Outstanding Graduate Honor and conducted research at the Institute of Intelligent Vehicles (TJU-IIV) directed by Prof. Yanjun Huang and Prof. Hong Chen.

My experience focuses on computer vision and robot learning. I am searching for full positions in 2025

🔥 News

  • 2024.12:  👏 Our V2XPnP paper on cooperative perception and prediction has been released. Enjoy it!
  • 2024.05:  👏 Our unsupervised reinforcement learning papers on autonomous driving have been accepted by IEEE Transactions on Intelligent Transportation Systems (Top-ranked in ITS Filed, IF:8.2)!
  • 2024.01:  💻 Start my research assistant in UCLA, focusing on prediction model implementation and map generation.
  • 2024.01:  💻 Start my student project in Teton, focusing on implementation and test Yolo-based 3D human mesh recovery model.
  • 2023.06:  🎉 Graduated from Tongji University with Outstanding Graduate Honor!
  • 2023.03:  🎉 Received the DTU ASU (Autonomous Systems) Master offer with Full Tuition Fee Waiver (30 per year)!
  • 2022.10:  💻 Start my internship in Volvo Cars, focusing on the development and implementation of a car defect detection model.

📖 Educations

  • 2023.09 - 2025.06, M.S., Autonomous Systems, Technical University of Denmark (DTU), Copenhagen, Denmark (Full Tuition Fee Waiver)
  • 2018.09 - 2023.06, B.Eng., Automotive Engineering, Tongji University (Tongji), Shanghai, China (Outstanding Graduate,Top 1%)

📝 Publications

sym

V2XPnP: Vehicle-to-Everything Spatio-Temporal Fusion for Multi-Agent Perception and Prediction

Zewei Zhou, Hao Xiang, Zhaoliang Zheng, Seth Z Zhao, Mingyue Lei, Yun Zhang, Tianhui Cai, Xinyi Liu, Johnson Liu, Maheswari Bajji, Jacob Pham, Xin Xia, Zhiyu Huang, Bolei Zhou, Jiaqi Ma

arXiv, 2024 | Paper | Project |

  • We proposed V2XPnP, the first open-source V2X spatio-temporal fusion framework for cooperative perception and prediction, and V2XPnP Sequential Dataset, the first large-scale, real-world V2X sequential dataset featuring multiple agents and all V2X collaboration modes (VC, IC, V2V, I2I).
sym

Unsupervised Reinforcement Learning for Multi-Task Autonomous Driving: Expanding Skills and Cultivating Curiosity

Zhenyu Ma, Xinyi Liu, Yanjun Huang

IEEE Transactions on Intelligent Transportation Systems, 2024

Top-ranked in ITS Filed, IF:8.2 | Paper

  • We introduced an unsupervised reinforcement learning method, Improved Contrastive Intrinsic Control (CIC), to help autonomous driving learn general skills and improve the generalization in different driving tasks.

🔬 Projects

sym

Automatic car body defect detection system

Volvo Cars, Shanghai, China, 2022.08 - 2022.10

Internship

  • Trained a defect detection model using Yolo v5, which can sort car body defects into dents and scratches, sort car seat defects into five types, whose accuracy arrived at 89% and 84% respectively in the test sets
  • Programmed the back-end API based on Django, whose functions include user-information management, image sets and detection algorithms upload, defect detection and results display
sym

Indoor High Precision Positioning and Navigation System Based on VSLAM

Tongji University, Shanghai, China, 2020.04 - 2021.04

Student Innovatation Training Program (SITP), Team Leader

  • Built 3D point cloud map using 2 ORB-SLAM2 frames
  • Developed WeChat navigation applet based on Feng-map open source map

🎖 Honors and Awards

  • 2023   Tongji Outstanding Student (Top 1%)
  • 2022   Third Prize of Challenge Cup Academic Science and Technology Competition, China (2.2% of 22,000)
  • 2022   Silver Model in Formula Student, China (Ranking: 2/51)
  • 2022   Infineon Technologies Scholarship (Top 3%)
  • 2021 & 2020   Weichai Scholarship (Highest endowed scholarship in Tongji sponsored by Weichai Co., Ltd., Top 2%, twice)
  • 2019   National Scholarship, China (Highest scholarship awarded by the Chinese government, Top 0.2%)

💻 Experience