An Interpretable Deep Learning Framework for Material Perception and Classification from Multisensory Tactile Data

Source: arXiv cs.RO·2026-08-25 12:00Publish time unverified

arXiv:2608.21894v1 Announce Type: new Abstract: Human tactile perception relies on complex multisensory cues. Yet the relationship between tactile signals and perceptual representations remains poorly understood, limiting the integration of touch in digital environments and human-like robotic perception.