Yixiao (Caleb) Feng

signature

Hi there! I am Yixiao Feng. I am a research assistant & incoming PhD student at the Hong Kong University of Science and Technology (Guangzhou), supervised by Prof. Renjing Xu.

Previously, I received my M.EngSc degree in Robotics 🤖 at University of New South Wales (UNSW) in 2025, supervised by Dr. Jose Guivant. Prior to that, I obtained my B.Eng. degree from Beijing University of Chemical Technology (BUCT).

My primary research interest lies in Embodied AI, encompassing robot learning and dexterous manipulation 🦾. If you are interested in collaborating or discussing any topics related to robotics, feel free to reach me out!

                 

profile photo

Recent News

  • [2025/01] 🎓 I've graduated from UNSW and obtained my M.EngSc degree in Robotics.
  • [2024/01] 🎉 Our work of BM-Loc was accepted by ICRA 2024.

Publications

*: equal contribution; †: corresponding author(s)
dise Block-Map-Based Localization in Large-Scale Environment
Yixiao Feng*, Zhou Jiang*, Yongliang Shi†, Yunlong Feng, Xiangyu Chen, Hao Zhao, Guyue Zhou

ICRA 2024 Oral
[Paper] [Video]

We first proposed a localization method based on Block-Map (BM) for large scenes, and provide a method for generating BMs and corresponding map-switching strategies. We also proposed a factor graph optimization method with different marginalization strategies for two cases where the robot moves within the same map or traverses between different maps. The final results showed that our method achieved a per-frame computation time up to 150% faster compared to using a global map.

Education

University of New South Wales

M.EngSc in Robotics (with Excellence), supervised by Dr. Jose Guivant
Feb. 2023 - Jan. 2025

Beijing University of Chemical Technology

B.Eng. in Mechanical Design, Manufacturing and Its Automation
Sep. 2018 - Jun. 2022
Ex-president of SIE Robotics Club


Research Experiences

The Hong Kong University of Science and Technology (Guangzhou)

Research Assistant, supervised by Prof. Renjing Xu
Jun. 2025 - Present

Qiyuan Lab

Research Intern, co-advised by Dr. Yongliang Shi and Dr. Weining Lu
Dec. 2024 - May 2025

Institute for AI Industry Research, Tsinghua University

Research Intern, co-advised by Dr. Yongliang Shi and Prof. Guyue Zhou
ARX, DISCOVER Lab
Oct. 2021 - Dec. 2022

Selected Projects
dise Large-Scale Globally Consistent Mapping using Solid-state LiDAR-based Place Recognition
Yixiao Feng, Jose Guivant


• Improved ImMesh through factor graph optimization by introducing STD loop-closure factors, and proposed a globally consistent mapping framework.
• Collected a LiDAR-inertial dataset consisting of 6 sequences with challenging low-texture scenarios, such as long staircases.
• Achieved SOTA localization accuracy on our self-collected dataset, meanwhile demonstrating the effectiveness of the globally consistent mapping framework.

dise Obstacle Avoidance for Indoor Racing Unmanned Vehicles
Xiangyu Chen, Zhengxiao Han, Yixiao Feng, Guyue Zhou

• Designed a multi-sensor fusion localization algorithm utilizing particle filtering and graph optimization.
• Utilized A* algorithm for global path planning and Timed Elastic Band (TEB) algorithm for local path planning and successfully avoided all tested dynamic and static obstacles.

dise Home Service Robot
Yixiao Feng, Zhengxiao Han, Yang Yang, Yi Zhu, Zeyu Ren, Xiang Gao, Chaowei Wang, Kexin Song

RoboCup @Home 2021


• Implemented Cartographer algorithm for robot localization and mapping using a single line LiDAR and an IMU.
• Designed a human following algorithm using the point cloud centroid on the tracked target, and implemented Openpose to recognize specific gestures to start and stop following.
• Developed a vision-based robotic grasping system combining YOLO-v3 object detection with point cloud processing for precise object localization.



© Yixiao Feng. Last update: .