Yutian is dedicated to improve the spatial understanding ability of autonomous systems, which requires a combination of geometry and semantic information about the surroundings. His research experience on visual-inertial SLAM at the AirLAB focuses on improving the recognition of scene geometry in adversarial conditions, while his previous experience in Natural Language Processing (NLP) and semantic segmentation focuses on grasping and interpreting the semantic of scenes.
yutianch@andrew.cmu.edu