Bashar I Ahmad*, James Murphy*, Patrick M Langdon*, Robert Hardy**, Simon J Godsill*. * Engineering Department, University of Cambridge, Trumpington ...
DESTINATION INFERENCE USING BRIDGING DISTRIBUTIONS Bashar I Ahmad*, James Murphy*, Patrick M Langdon*, Robert Hardy**, Simon J Godsill* * Engineering Department, University of Cambridge, Trumpington Street, Cambridge, UK, CB2 IPZ ** Jaguar Land Rover, Whitely, Coventry, UK
{
j
j
Emails: bia23. m362.pmI24.s g30
}@cam.ac.uk.rhardy @j aguarlandrover.com
ABSTRACT
We propose a novel probabilistic inference approach that per mits predicting, well in advance, the intended destination of a pointing gesture aimed at selecting an icon on an in-vehicle interactive display. It models the partial 3D pointing track as a Markov bridge terminating at a nominal destination. The solution introduced leads to a low-complexity Kalman-filter type implementation and is applicable in other areas in which early detection of the destination of a tracked object is bene ficial. Data collected in an instrumented vehicle illustrate that the proposed technique can infer the intent notably early in the pointing gesture. This can drastically reduce the pointing task time and visual-cognitive-manual attention required. human computer interactions, intent in ference, Kalman filter, bridging distributions. Index Terms-
1. INTRODUCTION
Interactive displays such as touchscreens are becoming an in tegrated part of the modern vehicle environment due to their ability to present large quantities of data associated with In Vehicle Infotainment Systems (IVIS) [1, 2, 3]. They are also easy to use via instinctive pointing gestures. However, us ing such displays entails dedicating a considerable amount of attention that would otherwise be available for driving, with serious safety implications [4, 5]. Additionally, due to driv ing or road conditions the user input can be highly perturbed, leading to erroneous selections, which compromises the sys tem usability and results in further distractions. In this paper, we propose a Bayesian intent inference ap proach that allows prediction, early in the pointing gesture, of the intended destination on an in-vehicle interactive dis play. This can significantly reduce pointing time and effort. Here, the pointing track is modelled as one of several Markov bridges, each incorporating one of the possible destinations, e.g. selectable icons on a GUI displayed on a touchscreen. The path of the pointing finger, albeit random, must end at the intended destination, i.e. it follows a bridge distribution from its start point to the destination. By determining the likelihood This work is supported by Jaguar Land Rover (JLR), Whitley, UK.
978-1-4673-6997-8/15/$31.00 ©2015 IEEE
•.
0
0
0
�'">(.
0 o
0
0
ci">
