Matteo Toscani

From haptic exploration to object perception in 3D

This study investigates how 3D object geometry guides haptic exploration and how the information gathered influences perceptual decisions, finding that global and local geometry integrate into recognition but memory limitations affect accuracy.

Unsupervised learning of haptic material properties

This study uses an autoencoder to learn a compressed representation of vibratory patterns from haptic exploration, enabling material classification and revealing similarities to human tactile perception.