Optimistic Programming of Touch Interaction
YANG
LI and HAO LU, Google Research
HAIMO
ZHANG, National University of Singapore
http://dx.doi.org/10.1145/2631914
Touch-sensitive surfaces have become a predominant input medium
for computing devices. In particular, multitouch capability of these devices
has given rise to developing rich interaction vocabularies for “real” direct
manipulation of user interfaces. However, the richness and flexibility of touch
interaction often comes with significant complexity for programming these
behaviors. Particularly, finger touches, though intuitive, are imprecise and
lead to ambiguity. Touch input often involves coordinated movements of multiple
fingers as opposed to the single pointer of a traditional WIMP interface. It is
challenging in not only detecting the intended motion carried out by these
fingers but also in determining the target objects being manipulated due to
multiple focus points. Currently, developers often need to build touch
behaviors by dealing with raw touch events that is effort consuming and
error-prone. In this article, we present Touch, a tool that allows developers
to easily specify their desired touch behaviors by demonstrating them live on a
touch-sensitive device or selecting them from a list of common behaviors.
Developers can then integrate these touch behaviors into their application as
resources and via an API exposed by our runtime framework. The integrated tool support
enables developers to think and program optimistically about how these
touch interactions should behave, without worrying about underlying complexity
and technical details in detecting target behaviors and invoking application
logic. We discuss the design of several novel inference algorithms that
underlie these tool supports and evaluate them against a multitouch dataset
that we collected from end users. We also demonstrate the usefulness of our
system via an example application.
M.
RACHMATARRAMADHAN (G64120049), Undergraduate Student
of Computer Science FMIPA IPB
Touchsensitive
screens telah menjadi
media input yang banyak digunakan
dalam perangkat mobile, dengan
memungkinkan pengguna untuk memanipulasi objek pada interface dan memanipulasi langsung UI (User Interface) membawa touch-based
interaction lebih dekat dengan penggunanya. Walaupun begitu touch-based interaction masih
menimbulkan masalah yang mendasar dalam pemrograman sebuah interaksi, yaitu memahami
karakteristik gerakan dari sebuah sentuhan, titik fokus ketika menggunakan
beberapa jari dan penerapan tindakan dalam pengaplikasian input sentuhan.