Chengliang Liu <br/> 柳程亮

Chengliang Liu
柳程亮

UCAS & SIAT, CAS


I am an incoming Ph.D. student at the Department of Computing (COMP), The Hong Kong Polytechnic University (PolyU) , supervised by Prof. Qing Li, co-supervised by Prof. Wenqi Fan. Currently, I am a third-year M.Sc. student at University of Chinese Academy of Sciences (UCAS) (learn more about UCAS), and doing my research at Shenzhen Institute of Advanced Technology (SIAT) , Chinese Academy of Sciences (CAS) under the supervision of Prof. Xinyu Wu and Prof. Zhengkun Yi. Before that, I graduated from School of Mechanical and Electrical Engineering, Central South University (CSU, Project 985, Project 211) with a bachelor’s degree.

My research interest mainly includes Lagrge Language Models (LLM), Machine Learning, and Robotics.

🔥 News

  • 2024.01: One co-author paper is accepted by IEEE ICRA 2024! 🎉
  • 2023.10: I am awarded a China postgraduate national scholarship (Top 1.67%, 20, 000 CNY)! 🏆
  • 2023.05: One first-author paper is accepted by IEEE TIM (JCR Q1, IF-5.6)! 🎉
  • 2023.04: One first-author paper is accepted by IEEE RCAR 2023! 🎉
  • 2022.07: I am honored with the title of Outstanding Student Leader of UCAS (Top 5%)! 🏅
  • 2022.07: I am honored with the title of Three Good Student of UCAS (Top 15%)! 🏅
  • 2021.05: I am honored with the title of Outstanding Graduate of CSU (Top 4%)! 🏅
  • 2021.05: I am honored with the title of Outstanding Graduate of Hunan Province (Top 3.5%)! 🏅
  • 2021.03: I have joined SIAT, CAS as a research assistant and will spend my final undergraduate year here! 🔬
  • 2020.12: I am awarded a Shanhe Elite First-class Education Scholarship (10, 000 CNY)! 🏆
  • 2020.11: I am awarded a CSU Second-Class Excellent Student Scholarship (Top 15%)! 🏆
  • 2020.09: I acquire a qualification of postgraduate recommendation, and decide to pursue my master’s degree in UCAS! 👨‍🎓
  • 2019.12: I am awarded a Weiqiao Aluminum Electricity Scholarship (5, 000 CNY)! 🏆
  • 2019.11: I am awarded a CSU First-Class Excellent Student Scholarship (Top 5%)! 🏆
  • 2018.11: I am awarded a CSU Second-Class Excellent Student Scholarship (Top 15%)! 🏆

📝 Publications

📌 Pinned

IEEE TIM
sym

A Deep Learning Method Based on Triplet Network Using Self-Attention for Tactile Grasp Outcomes Prediction(First-author, JCR Q1, IF-5.6)

Chengliang Liu, Zhengkun Yi, Binhua Huang, Zhenning Zhou, Senlin Fang, Xiaoyu Li, Yupo Zhang, Xinyu Wu.

IEEE Transactions on Instrumentation and Measurement

  • Investigate contrastive learning using triplet loss in a supervised manner for grasp outcome prediction in a small Gelsight-based tactile dataset.
  • Integrate cross self-attention mechanisms with triplet net to exploit context information from different sensors.
  • Found that a self-attention module can work as a non-linear projection head for contrastive learning, which performs better than a simple MLP.
  • The experimental results demonstrate significant performance improvements in various metrics when using our framework, compared to the original method.
IEEE RCAR 2023
sym

A Self-supervised Contrastive Learning Method for Grasp Outcomes Prediction(First-author)

Chengliang Liu, Binhua Huang, Yiwen Liu, Yuanzhe Su, Ke Mai, Yupo Zhang, Zhengkun Yi, Xinyu Wu

2023 IEEE International Conference on Real-time Computing and Robotics

  • Explore and verify the transferability of widely-applied data augmentation techniques in the field of Computer Vision to Gelsight-based tactile datasets.
  • A grasp outcome prediction network was proposed based on momentum contrast with the maintenance of a large dictionary, which achieves 81.83% predictive accuracy with a single tactile sensor data.
  • The experimental results show that the proposed network achieved state-of-the-art performance compared to three other self-supervised methods.

🔎 Other

👨‍💻 Experience

sym

SIAT, CAS, Shenzhen, China

Research Assistant @ Intelligent Systems and Robot Learning Lab (ISRL)

2021.03 - 2021.08

Teaching Assistant @ Introduction to Robotics

2023.02 - 2023.06

🏆 Selected Awards

🏅 Selected Honors