From Fall 2023 to Spring 2024, I work on the DTTD project, which has started since 2021, focusing on the 6D pose estimation problems. DTTD 3.0, which adds the dataset of robot components into existing rigid bodies of daily objects, provides model the ability to estimate rotation and translation of non-rigid-body moving robots. Challenges include error caused by multiple types of sensors, inaccuracy in capturing non-rigid-body movements of robots, and the processing time of pose-estimation algorithms.
