Dexterity from Touch: Self-Supervised Pre-Training of Tactile Representations with Robotic Play
This addresses the problem of fine-grained manipulation for robotics, where vision-based methods fail due to contact forces or occlusions, offering a domain-specific improvement.
The paper tackles the challenge of teaching dexterity to multi-fingered robots by introducing T-Dex, a tactile-based approach that uses self-supervised pre-training on play data and non-parametric policies, resulting in models that outperform purely vision and torque-based models by an average of 1.7X across five tasks.
Teaching dexterity to multi-fingered robots has been a longstanding challenge in robotics. Most prominent work in this area focuses on learning controllers or policies that either operate on visual observations or state estimates derived from vision. However, such methods perform poorly on fine-grained manipulation tasks that require reasoning about contact forces or about objects occluded by the hand itself. In this work, we present T-Dex, a new approach for tactile-based dexterity, that operates in two phases. In the first phase, we collect 2.5 hours of play data, which is used to train self-supervised tactile encoders. This is necessary to bring high-dimensional tactile readings to a lower-dimensional embedding. In the second phase, given a handful of demonstrations for a dexterous task, we learn non-parametric policies that combine the tactile observations with visual ones. Across five challenging dexterous tasks, we show that our tactile-based dexterity models outperform purely vision and torque-based models by an average of 1.7X. Finally, we provide a detailed analysis on factors critical to T-Dex including the importance of play data, architectures, and representation learning.