An Experimental Study of Visual and Haptic Object Recognition

概要

論文の詳細を見る
Object recognition was investigated in within and cross-modal matching experiments. Three-dimensional nonsense objects were learned (observed) and tested (recognized) under four conditions:(1) visual learning and visual testing;(2) haptic learning and haptic testing;(3) visual learning and haptic testing; and (4) haptic learning and visual testing. Results indicated that recognition as measured by the mean errors was best in the visual-visual condition corroborating past findings. Moreover, the performance in the other within-modal haptic-haptic condition was superior to the cross-modal, haptic-visual condition. The enhanced performance of within-modal matching is discussed in relation to the idea of viewpoint dependency in visual and haptic object recognition.
昭和女子大学の論文
2006-03-31