Xinshuo's research interests lie in seamlessly integrating modules across Autonomous Vehicle (AV) stacks by building next-generation differentiable AV pipelines, propagating uncertainties from the upstream to downstream modules to improve end-to-end performance, and exploring the most effective structure of differentiable pipelines. Her past research experience in AV spans the tasks of object detection, multi-object tracking, re-identification, trajectory prediction, and end-to-end vision-based planning. More information about Xinshuo can be found here: https://xinshuoweng.com/
