cv
This is the CV for a modern roboticist.
Basics
Name | Yuan Shenghai |
Label | Scientist |
shyuan AT ntu.edu.sg | |
Phone | +65 XXXX XXXX |
Url | https://snakehaihai.github.io/publications/ |
Summary | A Chinese-born Singaporean robotics researcher and expert, combining extensive field experience with a strong background in cutting-edge robotics research. |
Work
-
2019.11 - Now Senior Research Fellow
Nanyang Technological University
Conduct research activity, writing papers, write project proposals, finish projects, writing research proposals, raising PHD kid
- Robotics
Education
-
2014.01 - 2019.01 Singapore
Doctor of Philosophy
Nanyang Technological University, Singapore
Robotics
- Electrical and Electronic Engineering
-
2010.01 - 2014.01 Singapore
Bachelor degree
Nanyang Technological University, Singapore
Robotics
- Electrical and Electronic Engineering
Awards
- 2014.1.01
NTU Research Scholarship
Nanyang Technological University, Singapore
The NTU Research Scholarship is awarded to outstanding graduate students for research.
- 2024.5.10
ICRA Conference Editorial Board Awards
Nanyang Technological University, Singapore
ICRA 2024 Conference Editorial Board Outstanding Reviewers
- 2023.1.01
Third place in the visual-inertial track of the 2023 ICCV SLAM Challenge.
ICCV 2023 SLAM Challenge
- 2021.1.01
Second place in the academic track of the 2021 Hilti SLAM Challenge.
Hilti SLAM Challenge
- 2013.1.01
Champion out of 26 teams in Singapore DSTA Techx Swarm Challenge.
Defence Science and Technology Agency (DSTA), Singapore
- 2012.1.01
Ranked 6 Out of 140 teams in DAPRA UAVforge Challenge 2012.
Defense Advanced Research Projects Agency (DAPRA), Unite States
- 2011.1.01
Champion out of 35 teams in 2011 Taiwan UAV Reconnaissance Challenges.
National Cheng Kung University
Volunteer
-
2023.12 - 2025.12 Reviewer
ICRA, IROS, TRO, RAL, CVPR, NeurIPS, TCDS, ISPRS, TIE, etc.
I critically evaluate a manuscript's quality, originality, and validity, provide constructive feedback, and recommend a decision while maintaining confidentiality and ethical standards.
- Received Best Reviewer award from IEEE.
-
2023.12 - 2025.12 Associate Editor
Unmanned Systems
Manage and oversee the peer review process to ensure it is fair, rigorous, and timely. Maintain and improve the quality of content by ensuring publications meet high academic and professional standards.
- Control the review process to be 90% within 2 months for each round of review.
Publications
-
2024.12.01 Robust Loop Closure by Textual Cues in Challenging Environments
IEEE Robotics and Automation Letters
This study proposes a multi-modal loop closure method that leverages explicit human-readable textual cues to enhance robot navigation in featureless, degenerative, and repetitive environments.
-
2024.01.01 Graph Optimality-Aware Stochastic LiDAR Bundle Adjustment with Progressive Spatial Smoothing
arXiv preprint arXiv:2410.14565
This study introduces PSS-GOSO, a robust, efficient, and scalable LiDAR Bundle Adjustment method. It features Progressive Spatial Smoothing for robust LiDAR feature association and Graph Optimality-aware Stochastic Optimization for efficient optimization, validated across diverse scenes for superior performance.
-
2024.01.01 AirSLAM: An Efficient and Illumination-Robust Point-Line Visual SLAM System
arXiv preprint arXiv:2408.03520
AirSLAM is a visual SLAM system designed to handle both short-term and long-term illumination challenges. It combines deep learning techniques for feature detection and matching with traditional backend optimization methods, utilizing a unified convolutional neural network to extract keypoints and structural lines for improved accuracy and robustness.
-
2024.01.01 Incremental Joint Learning of Depth, Pose and Implicit Scene Representation on Monocular Camera in Large-Scale Scenes
arXiv preprint arXiv:2404.06050
This paper proposes an incremental joint learning framework that achieves accurate depth and pose estimation, along with large-scale scene reconstruction, using a monocular camera. The method employs a vision transformer-based network for enhanced scale information estimation and introduces a feature-metric bundle adjustment for robust camera tracking.
-
2024.01.01 Compact 3D Gaussian Splatting for Dense Visual SLAM
arXiv preprint arXiv:2403.11247
This research presents a method to reduce memory and storage costs in 3D Gaussian-based SLAM systems while maintaining high-quality scene reconstruction and rendering, utilizing techniques like sliding window-based masking and codebook-based compression.
-
2024.01.01 GPTR: Gaussian Process Trajectory Representation for Continuous-Time Motion Estimation
arXiv preprint arXiv:2410.22931
This paper introduces a Gaussian Process Trajectory Representation (GPTR) framework for continuous-time motion estimation, employing a third-order random jerk model to provide smooth, continuous trajectory representations crucial for precise estimation of complex motion.
Skills
Physics | |
ROS | |
C++ | |
Python | |
Machine Learning | |
Computer Vision |
Languages
Chinese | |
Native speaker |
English | |
Fluent |
Interests
Robotics | |
Unmanned Aerial Vehicles | |
Simultaneous Localization And Mapping, | |
Computer Vision | |
Point Cloud Registration | |
Audio Data | |
Benchmarking |
References
Professor Xie Lihua | |
2012-Now. Email: elhxie AT ntu.edu.sg. Prof. Xie, a distinguished Fellow of IEEE, is widely recognized for his outstanding contributions to the field of engineering and technology. As my postdoc supervisor, he has been an exceptional mentor, offering invaluable guidance and support throughout numerous projects and events. His expertise and leadership have profoundly influenced my academic and professional journey. Our collaboration began in 2012, during the prestigious Million Dollar DSTA TechX Challenge, where I joined the NTU team under his mentorship. That experience not only marked the start of our professional relationship but also set the foundation for many successful endeavors together. |
Dr. Thien-Minh Nguyen | |
2017-Now. Email: thienminh.npn AT ieee.org. Dr. Thien Minh is a Research Assistant Professor at Nanyang Technological University (NTU). Over the years, we have collaborated extensively, publishing numerous high-impact research papers together. Beyond being a trusted colleague, Thien Minh is also a close friend, and we often engage in productive brainstorming sessions to tackle challenging problems. Whether it’s reviewing papers, writing code, or generating ideas for new projects, we frequently rely on each other’s expertise and perspectives to refine our work and ensure the highest standards of quality and innovation.. |
Professor Wang Han | |
2012-2019. Email: HW AT ntu.edu.sg. Prof. Wang Han, my PhD supervisor, has played a pivotal role in shaping my academic journey. He has a remarkable ability to identify key research problems and has guided me through a series of impactful projects, offering me opportunities to step onto larger stages. Our connection goes back to my undergraduate years, where he first began teaching me. Our collaborative journey truly began in 2012, during the prestigious Million Dollar DSTA TechX Challenge, which laid the foundation for many achievements that followed. He is now in Xiamen university Malaysia. |