Multimodal Visual-Tactile Representation Learning through Self-Supervised Contrastive Pre-Training
Vedant Dave,Fotios Lygerakis,Elmar Rueckert,Vedant Dave,Fotios Lygerakis,Elmar Rueckert
The rapidly evolving field of robotics necessitates methods that can facilitate the fusion of multiple modalities. Specifically, when it comes to interacting with tangible objects, effectively combining visual and tactile sensory data is key to understanding and navigating the complex dynamics of the physical world, enabling a more nuanced and adaptable response to changing environments. Neverthel...
