hide
Free keywords:
-
Abstract:
When integrating estimates from redundant sensory signals, humans seem to weight these estimates according to their reliabilities. In the present study, human observers used active touch to judge the curvature of a shape. The curvature was specified by positional and force signals: When a finger slides across a surface, the fingers position follows the surface geometry (position signal). At the same time it is exposed to patterns of forces depending on the gradient of the surface (force signal; Robles-de-la Torre amp; Hayward, 2001). We show that variations in the surfaces material properties (compliance, friction) influence the sensorily available position and force signals, as well as the the noise associated with these signals. Along with this, material properties affect the weights given to the position and force signals for curvature judgements. Our findings are consistent with the notion of an observer who weights signal estimates according to their reliabilities. That is, signal
wei
ghts shifted with the signal noise, which in the present case resulted from active exploration.