Jui-Te "Ray" Huang

mmWave radar enables robust ego-velocity estimation in visually denied conditions, but existing methods rely on sparse point clouds prone to limited FOV and Doppler aliasing. UNRIO instead uses raw I/Q signals with a radar foundation model, combining uncertainty-aware velocity estimates with IMU measurements for state estimation. It demonstrates robust performance across diverse scenarios, including lateral motion and severe Doppler aliasing.

We propose a SAR-based pipeline for probabilistic occupancy mapping using mmWave radar in challenging conditions like smoke and fog. Extensive experiments evaluate mapping accuracy, path planning, and system parameters. We also release an open-source radar dataset and GPU-accelerated processing pipeline.

In this paper, we address Doppler velocity measurement uncertainties. We present a method to optimize body frame velocity while managing Doppler velocity uncertainty. Based on our observations, we propose a dual imaging radar configuration to mitigate the challenge of discrepancy in radar data. To attain high-precision 3D state estimation, we introduce a strategy that seamlessly integrates radar data with a consumer-grade IMU sensor using fixed-lag smoothing optimization. Finally, we evaluate our approach using real-world 3D motion data.

We introduce the CRUW3D dataset, including 66K synchronized and well-calibrated camera, radar, and LiDAR frames in various driving scenarios. Unlike other large-scale autonomous driving datasets, our radar data is in the format of radio frequency (RF) tensors that contain not only 3D location information but also spatio-temporal semantic information.

We propose the use of single-chip millimeter-wave (mmWave) radar, which is lightweight and inexpensive, for learning-based autonomous navigation. However, because mmWave radar signals are often noisy and sparse, we propose a cross-modal contrastive learning for representation (CM-CLR) method that maximizes the agreement between mmWave radar data and LiDAR data in the training stage to enable autonomous navigation using radar signal.

The DARPA Subterranean (SubT) Challenge aims to develop innovative technologies that would augment operations underground. The SubT Challenge will explore new approaches to rapidly map, navigate, search, and exploit complex underground environments. I participated in the Urban Circuit as the software team lead and human supervisor. We built our robots to perform search and rescue missions in an unfinished nuclear power plant.

Facilitating navigation in pedestrian environments is critical for enabling people who are blind and visually impaired (BVI) to achieve independent mobility. A deep-reinforcement-learning-based assistive guiding robot with ultrawide-bandwidth (UWB) beacons that can navigate through routes with designated waypoints was designed in this study.

This project is to help people who are interested in marine robotics. We built a low-cost surface vehicle called duckieboat. Our vehicles are compatible with machine learning frameworks. We demonstrate the implementation of classic autonomous navigation algorithms and tracking vehicles using deep learning computer vision and sensors on board.