The field of robotics is witnessing significant advancements in dexterous manipulation, with a focus on achieving human-like capabilities. Recent research has emphasized the importance of tactile sensing, reinforcement learning, and imitation learning in developing robots that can perform complex tasks. The integration of vision, language, and tactile sensing is also becoming increasingly prominent, enabling robots to better understand their environment and interact with objects. Furthermore, advances in soft robotics and continuum robots are expanding the possibilities for versatile manipulation. Noteworthy papers in this area include the development of Tactile-VLA, a framework that combines vision, language, action, and tactile sensing, and the introduction of Octopi-1.5, a visual-tactile-language model that can process tactile signals from multiple object parts. Additionally, the Robot Drummer system has demonstrated the potential of reinforcement learning in teaching humanoid robots to perform expressive drumming tasks.
Advances in Dexterous Robotic Manipulation
Sources
Flexible arrangement of two Bennett tubes
Robotic Calibration Based on Haptic Feedback Improves Sim-to-Real Transfer
Towards Human-level Dexterity via Robot Learning
Tactile-VLA: Unlocking Vision-Language-Action Model's Physical Knowledge for Tactile Generalization
Demonstrating the Octopi-1.5 Visual-Tactile-Language Model
Hand Gesture Recognition for Collaborative Robots Using Lightweight Deep Learning in Real-Time Robotic Systems
Simulations and experiments with assemblies of fiber-reinforced soft actuators
Exteroception through Proprioception Sensing through Improved Contact Modeling for Soft Growing Robots
Unified Modeling and Structural Optimization of Multi-magnet Embedded Soft Continuum Robots for Enhanced Kinematic Performances
Closed Form Time Derivatives of the Equations of Motion of Rigid Body Systems
Robot Drummer: Learning Rhythmic Skills for Humanoid Drumming
The Developments and Challenges towards Dexterous and Embodied Robotic Manipulation: A Survey
Keep the beat going: Automatic drum transcription with momentum
Few-shot transfer of tool-use skills using human demonstrations with proximity and tactile sensing