Semantic Pose
Semantic pose research focuses on understanding and representing the pose of objects or humans in a way that incorporates meaningful contextual information, going beyond simple geometric location. Current research emphasizes robust pose estimation using deep learning models, often incorporating keypoint detection and convolutional neural networks, with a focus on improving accuracy and robustness against factors like motion blur, varying lighting, and semantic perturbations. This work has significant implications for applications ranging from sign language recognition and athletic performance analysis to augmented reality and visual localization systems, improving the accuracy and reliability of these technologies.
Papers
August 17, 2024
July 31, 2024
September 2, 2023
May 17, 2023
November 12, 2022
April 12, 2022
March 31, 2022