Ruohai Ge

PhD student in Robotics

University of Southern California
Los Angeles, U.S.

/ / /

About me

I am a second-year Ph.D. student in Computer Science at the University of Southern California, working with Prof. Gaurav S. Sukhatme in the Robotic Embedded Systems Laboratory (RESL). My research focuses on the intersection of heterogeneous multi-robot systems and robot learning, aiming to develop intelligent, collaborative behaviors across diverse robotic platforms.

I earned my M.S. degree in Robotics from Carnegie Mellon University, where I was advised by Prof. Sebastian Scherer in the AirLab. Before that, I completed a B.S. in Electrical and Computer Engineering with a major in Robotics at CMU, working in the Biorobotics Lab under Prof. Howie Choset.

Publications

*: Equal contribution.

iLoc: An Adaptive, Efficient, and Robust Visual Localization System
Peng Yin*, Shiqi Zhao*, Jing Wang*, Ruohai Ge, Jianmin Ji, Yeping Hu, Huaping Liu, Jiandan Han
IEEE Transactions on Robotics (T-RO), 2025
An innovative visual localization system designed to enhance the autonomy and adaptability of robotic agents in long-term and large-scale applications.

Automerge: A framework for map assembling and smoothing in city-scale environments
Peng Yin, Shiqi Zhao, Haowen Lai, Ruohai Ge, Ji Zhang, Howie Choset, Sebastian Scherer
IEEE Transactions on Robotics (T-RO), 2023
A framework for assembling LiDAR map segments into a complete map.

Projects

Real-Time Visual Localization System in Changing and Challenging Environments via Visual Place Recognition
Master's Thesis
Feb 2021 - Aug 2022
Report Presentation Demo1 Demo2
This system excels in real-time visual localization, handling challenges like illumination changes, dynamic objects, and random starting points.

ALITA: A Large-scale Incremental Dataset for Long-term Autonomy
Public Dataset
Feb 2021 - Aug 2022
Project Page Manuscript Dataset
A long-term place recognition dataset for mobile localization in dynamic environments, including campus and city tracks. The campus track is recorded on 10 trajectories. Each is recorded 8 times under variant illumination conditions. The city track is recorded on 120km trajectory.

Autonomous Trash Can
18-500 ECE Design Experience (Bachelor Capstone Project)
Sep 2019 - Dec 2019
Report Demo Code
Built a trash can that can automatically catch thrown trash through computer vision and a mobile robot.

Brachiation Robot
BioRobotics Lab, Carnegie Mellon University
May 2019 - Sep 2019
Manuscript Demo
Assisted in design and implementation of a three-Link brachiation robot with optimal control based trajectory tracking controller.

Embedded Visual Processing
BioRobotics Lab, Carnegie Mellon University
Jun 2019 - Dec 2019
Fastener Classification Laser Extraction
Did image classification, feature detection and laser extraction on fasteners on an embedded platform (i.MX RT1064).

Robotic Block Builder
CMU 16-264 Humanoids
Feb 2019 - May 2019
Demo Simulation Code
Built a robotic arm for building and adjusting blocks with AI reasoning and computer vision.

Salamander Gait Tracking
BioRobotics Lab, Carnegie Mellon University
Jun 2018 - August 2018
All Feet Body
Tracked key points of a moving salamander through videos to design and compare gaits for a quadruped robot with a real salamander.

Services

Reviewer

IJRA

Workshop Organization

Co-organizer of Workshop Competition - Closing the Loop on Localization, IROS 2023
Co-organizer of General Place Recognition Competition, ICRA 2022

Teaching

18-500 ECE Design Experience , TA, CMU, 2020
16-311 Introduction to Robotics , TA, CMU, 2020
16-311 Introduction to Robotics , TA, CMU, 2019
82-242 Intermediate Spanish II , TA, CMU, 2018