标题: A haptically-augmented interface for digital character animation [打印本页] 作者: 彬彬 时间: 2011-12-30 18:52 标题: A haptically-augmented interface for digital character animation 1 Introduction
The majority of the commercially available software are accessed
through the Windows/Icon/Mouse/Pointer (WIMP) interfaces. Arguably,
these interfaces fail to communicate the richness and complexity
of human gesture [Scali et al. 2003]. Consequently, embodied
skills such as those of creative makers cannot be easily accommodated
in digital workspaces. In [Dima et al. 2010] we illustrated
this issue in the traditional technique of puppet Stop-motion Animation
and, as a first step, explored how puppet stop-motion animators
perceive a virtual workspace mounted with the haptic modality and
enabling gestural action. The investigation was driven by the development
of a new interface system for animating digital 3-D articulated
characters designed in collaboration with puppet Stop-motion
animators. Hereby, we present further insights on the perception of
the virtual workspace derived from the design study on bi-manual
interaction. The system itself contributes a new interaction mechanism
in the field of Human-Computer Interaction.
2 System Design
The virtual workspace was a 2 1=2D space on the computer screen
mounted with a set of backgrounds and models which could be selected
from a drop down menu. A skeleton (or rig), a hierarchical
chain of bones and joints, was attached to each virtual character
making its movement possible via the implementation of kinematic
algorithms. An on-screen animation controller provided information
about each captured frame and keyboard buttons were used to
set/advance/retrace/swap and delete a frame as well as save or load
a sequence of frames.
One 6-DoF stylus-based haptic device, Sensable Omni, was used
to manipulate the whole or parts of the virtual character’s skeleton
as well as the lights and cameras. The device was represented on
screen with a sphere which could be moved in three dimensions
within the virtual workspace. For the bi-manual interaction study
we used two input devices for the actions of the non-dominant hand
which were initially assigned to moving the camera: the Wii controller
and Connexxion’sTM3D mouse . We tested various setups
for the Wii. In the most preferable one, the animators would hold
the Wii vertically to the ground. The Wii transfered mimetically to
the camera viewpoint the animators’ gestures which corresponded
to tilting, spinning and rolling, moving up and down. Moving back
e-mail: marizadima@ed.ac.uk
ye-mail:mark.wright@ed.ac.uk
ze-mail:john.lee@ed.ac.uk
and forth, right and left was done by pressing the cross buttons. By
pressing the back button the camera would stop moving. A keyboard
button positioned the camera in close distance to the character
and keeping button ’A’ on the Wii pressed would put it in orbit
around it. Using this setup, we asked a team of three stop-motion
animators to manipulate and animate an articulated 3-D character
with no specific time limitations in order to explore how they
perceived the virtual environment through haptically-augmented bimanual
work.
3 Results
In [Dima et al. 2010] we discussed the difficulty the animators had
in perceiving the virtual space in relation to how they moved the
haptic device’s representation in it. We explored the solution of
providing visual cues (e.g. visual markers on the selected bones)
which led to the observation that visual and haptic cues provide
greater perception of space and ease its navigation. The bi-manual
interaction study revealed ways of improving their perception of the
space kinaesthetically.
Although the Wii was considered to be a less static device in comparison
to the 3D mouse, the animators would prefer neither of the
two for camera manipulation. They suggested that this task could
be carried out with the haptic controller in the dominant hand using
the device’s buttons on the stylus to switch between camera navigation
and character manipulation. The controllers, and particularly
the Wii, could instead be used to move or rotate the whole character
while the dominant hand would use the haptic device to perform
the subtle manipulation of the character’s skeleton parts. We found
that this method improved their perception of the virtual space as
opposed to having both hands working in different things without
being co-located, even roughly, on the virtual space. In addition, the
animators occasionally felt that the sense of touch that the device
recreated assisted them in navigating around the skeleton’s structure
and in understanding the position of the device’s representation
during navigation.
References
DIMA, M., WRIGHT, M., AND LEE, J. 2010. Exploration of direct
bi-manual interaction in digitally mediated stop-motion animation.
In Proceedings of NordiCHI 2010, NordiCHI 2010.
SCALI, S., WRIGHT, M., AND SHILLITO, A. M. 2003. 3d modelling
is not for wimps. In Proceedings of HCI International,