Tactile Estimation

Tactile estimation focuses on developing methods for robots to accurately perceive and interpret tactile information, enabling more dexterous and adaptable manipulation. Current research emphasizes integrating tactile data with visual information, using various approaches including factor graphs, dual filtering, and neural implicit representations, to improve object property estimation (e.g., mass, stiffness, shape) and control during tasks like grasping, placement, and in-hand manipulation. This work is significant because accurate tactile perception is crucial for robots to interact effectively with unstructured environments and perform complex manipulation tasks requiring fine motor skills, ultimately advancing robotics capabilities in areas such as manufacturing and assistive technologies.