Christina Dao Wen Lee

Christina Dao Wen Lee


Research Focus & Metrics

Academic Career Length

5Years (2022-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
9 January 2023 in IEEE Robotics and Automation Letters
Verified
Journal Article
Closed access
ZLZhikai Li
Zhikai Li
CLChristina Dao Wen Lee
Christina Dao Wen Lee
BTBeatrix Xue Lin Tung
Beatrix Xue Lin Tung
ZHZefan Huang
Zefan Huang
DRDaniela Rus
Daniela Rus
Marcelo H.   Ang Jr.Marcelo H. Ang Jr.
Marcelo H. Ang Jr.
Hot-NetVLAD implements a hot-spot detector on a learned local key-patch descriptor algorithm for Visual Place Recognition (VPR), thereby greatly cutting down the size of features extracted. The hot-spots pinpoint which regions are crucial for comparison when performing VPR. As hot-spots land on only a small portion of the feature space, the number of local descriptors extracted is greatly reduced. A novel method to extract ground truths of hot-spots in the context of VPR is proposed so that the hot-spot detector in Hot-NetVLAD can be trained for VPR purposes. Hot-NetVLAD is evaluated on the Pittsburgh250k and Tokyo24/7 datasets. While results show that Hot-NetVLAD trades some accuracy loss for storage efficiency, the recall remains competitive when compared to state-of-the-art methods. Furthermore, identified hot-spots bring new insights to key regions required for VPR, as they tend to fall on distinguishable static objects in the scene. This can potentially be applied to increase the robustness of mobile robot localization by increasing resilience to dynamic environments, whilst still being able to perform static obstacle matching effectively.

18 November 2024 in IEEE Robotics and Automation Letters
Verified
Journal Article
Closed access
MLMin Young Lee
Min Young Lee
CLChristina Dao Wen Lee
Christina Dao Wen Lee
JLJianghao Li
Jianghao Li
Marcelo H.   Ang Jr.Marcelo H. Ang Jr.
Marcelo H. Ang Jr.
In the advancing domain of autonomous driving, this research focuses on enhancing 3D Multi-Object Tracking (3D-MOT). Pedestrians are particularly vulnerable in urban environments, and robust tracking methodologies are required to understand their movements. Prevalent Tracking-By-Detection (TBD) frameworks often underutilize the rich visual data from sensors such as cameras. This study leverages the advanced visual foundation model, DINOv2, to refine the TBD framework by incorporating camera modality, thereby improving pedestrian tracking consistency and overall 3D-MOT performance. The proposed DINO-MOT framework is the first application of DINOv2 for enhancing 3D-MOT through pedestrian Re-Identification (Re-ID), and Score Filter Ceiling is implemented to prevent premature exclusion of low-confidence 3D detections during tracking association. Furthermore, utilization of DINOv2 as a feature extractor within the DINO-MOT framework reduces pedestrian ID switches by up to 12.3%. Achieving AMOTA of 76.3% on the nuScenes test dataset, DINO-MOT has set a new benchmark in the 3D MOT literature with an improvement of 0.5%, securing the top rank on the leaderboard. Furthermore, this research paves the potential of applying a visual foundation model to improve the existing TBD framework, to enhance 3D-MOT in autonomous driving.

11 March 2022 in Machines
Verified
Journal Article
Open access
PDF file
YCYue Linn Chong
Yue Linn Chong
CLChristina Dao Wen Lee
Christina Dao Wen Lee
LCLiushifeng Chen
Liushifeng Chen
CSChongjiang Shen
Chongjiang Shen
KCKen Kok Hoe Chan
Ken Kok Hoe Chan
Marcelo H.   Ang Jr.Marcelo H. Ang Jr.
Marcelo H. Ang Jr.
We tackle the problem of achieving real-world autonomous driving for buses, where the task is to perceive nearby obstacles and predict their motion in the time ahead, given current and past information of the object. In this paper, we present the development of a modular pipeline for the long-term prediction of dynamic obstacles’ trajectories for an autonomous bus. The pipeline consists of three main tasks, which are the obstacle detection task, tracking task, and trajectory prediction task. Unlike most of the existing literature that performs experiments in the laboratory, our pipeline’s modules are dependent on the introductory modules in the pipeline—it uses the output of previous modules. This best emulates real-world autonomous driving and reflects the errors that may accumulate and cascade from previous modules with less than 100% accuracy. For the trajectory prediction task, we propose a training method to improve the module’s performance and attain a run-time of 10 Hz. We present the practical problems that arise from realising ready-to-deploy autonomous buses and propose methods to overcome these problems for each task. Our Singapore autonomous bus (SGAB) dataset evaluated the pipeline’s performance. The dataset is publicly available online.

1 January 2026 in IEEE Transactions on Intelligent Transportation Systems
Verified
Journal Article
Closed access
XHXunjie He
Xunjie He
CLChristina Dao Wen Lee
Christina Dao Wen Lee
MWMeiling Wang
Meiling Wang
CYChengran Yuan
Chengran Yuan
ZHZefan Huang
Zefan Huang
Marcelo H.   Ang Jr.Marcelo H. Ang Jr.
Marcelo H. Ang Jr.