Guan-Horng Liu

We don’t have enough information about this author to calculate their statistics. If you think this is an error let us know.
Learn More
Motion planning for unmanned ground vehicle on rough terrain has been an active research topic for a long time. One of the main challenges comes from the complexity of the off-road field environment. In this project, we aimed to propose a motion planner for a full-size all-terrain vehicle (ATV) for the application of off-road autonomous navigation. More(More)
Sensor fusion is indispensable to improve accuracy and robustness in an autonomous navigation setting. However, in the space of end-to-end sensorimotor control, this multimodal outlook has received limited attention. In this work, we propose a novel stochastic regularization technique, called Sensor Dropout, to robustify multimodal sensor policy learning(More)
Recently, reinforcement learning with deep neural networks has achieved great success in challenging continuous control problems such as 3D locomotion and robotic manipulation. However, in real-world control problems, the actions one can take are bounded by physical constraints, which introduces a bias when the standard Gaussian distribution is used as the(More)
In this thesis, we focus on making robots more trustable by making them describe and explain their actions. First, to tackle the problem of making robots describe their experience, we introduce the concept of verbalization, a parallel to visualization. Our verbalization algorithm can analyze log files as well as the robot’s live execution data to produce(More)
This thesis explores both traditional motion planning and end-to-end learning algorithms in the off-road settings. We summarize the main contributions as 1) propose an RRT-based local planner for high-speed maneuvering, 2) derive a novel stochastic regularization technique that robustifies end-to-end learning in the spirit of sensor fusion, and 3)(More)
  • 1