ANIMATE ARTS SENIOR PROJECT, MIDWAY UPDATE - PowerPoint PPT Presentation

1 / 13
About This Presentation
Title:

ANIMATE ARTS SENIOR PROJECT, MIDWAY UPDATE

Description:

Accuracy is improving, though not perfect. Demo on Wiki for the white ... Code more or less ready to go, still needs perfecting. On Schedule, Winter quarter ... – PowerPoint PPT presentation

Number of Views:40
Avg rating:3.0/5.0
Slides: 14
Provided by: dadaCsNor
Category:

less

Transcript and Presenter's Notes

Title: ANIMATE ARTS SENIOR PROJECT, MIDWAY UPDATE


1
ANIMATE ARTS SENIOR PROJECT, MIDWAY UPDATE
2
CONCEPT
An interactive installation in which viewers
attention exerts selective force on perpetually
recombining graphic elements. The basic
hardware consists of a digital projector, a
computer, and a motion camera. Genetic
algorithms breed and draw discrete graphic
entities on the screen. Image processing
software uses camera data to establish position
of viewers relative to the screen. If the
position of a viewer matches the position of an
entity, its fitness will increase meaning that
its aesthetic properties will endure in future
generations.
3
ORIGINAL SCHEDULE
Midway Objectives Sort out what is and isnt
possible image-processing wise. Solve technical
dilemmas. Funding? Finish majority of coding,
infrastructure, to allow focus on aesthetic
decisions and collection of graphic primitives
Spring quarter.
4
WHAT WENT RIGHT
Simple face tracking is possible in
Flash. Therefore, it should be feasible to
execute the project in Flash allowing for
access to the Flash drawing and animation
libraries I know and love. Face tracking
addresses technical dilemma of collecting user
input. Schedule still partially intact Ready
to move to rendering and recombination portions
using Flash should speed this up
5
WHAT WENT WRONG
Face tracking isnt eye tracking. Lots of time
spent exploring technical options that most
likely wont make final iterations of the
project. The Wiimote DirectX and other
graphics APIs Fiddling with animation systems
attempting to reimplement Flash
functionality Learning elementary C Syntax
was familiar, but the Windows APIs were brutal
6
WHAT I WOULD HAVE DONE DIFFERENTLY
Though it sounds counterintuitive Less planning
research, more coding design Theres a
planning saturation point Research as a form of
creative avoidance Excessive researchs happy
externality Selected links on the Wiki
7
STATUS
Face tracking working Supports multiple
faces Supports distance estimation Accuracy is
improving, though not perfect Demo on Wiki for
the white wall / web cam endowed Research into
genetic programming under way, some crude tests
completed in C, considering options in Flash GA
vs. GP? Shifting code to Flash 9 / ActionScript
3 Rewrite of the AS virtual machine 10x
Faster! according to Adobe Fast enough?
8
WHAT HAPPENED TO EYE TRACKING?
As I was told before I started, eye tracking
has Terminal technical and logistical
problems Huge processing and hardware
demands Small area without elaborate panning
mechanisms Very sketchy multi-user
support Demands on the user Calibration Fixe
d position or Unwelcome nerdery / physical
imposition Video camera glasses, arrow hats,
etc. Allows immediate reverse engineering of
situation
9
WHY FACE TRACKING?
Lighter hardware and processing
demands Multi-user support Additive
occlusion Address precision-related pitfalls
with installation design e.g. A choice of
screens / More foot traffic / Back vs Front
of Head
10
DOES FACE TRACKING CHANGE CONTENT?
Less about vision (though still a
component) Position represents a decision,
captured by the system and reflected visually by
the rendering code as selective force. Face
position is a less precise measure of vision than
eye position. New system favors larger number of
viewers. Still about aggregation of
preference Consensus without communication Still
about creative machines Implications of graphic
recombination unchanged More about human /
machine interaction Physical engagement
increased
11
THE CURRENT (REVISED) SCHEDULE
Still three (possibly four) basic fronts to the
project 1. Detection Code more or less ready to
go, still needs perfecting On Schedule, Winter
quarter 2. Recombination Genetic code
researched but not written. Winter quarter ?
Spring quarter 3. Rendering Animation system
will extend Flash. On schedule, Spring
quarter 4. Acoustics Dynamic sound to reflect
events on the screen. AS3 has improved sound
support. Spring quarter
12
GENERAL ADVICE
Embrace the kludge Even if it shouldnt work, it
might Avoid feature fixation Projects
evolve Take lt 4 courses if possible Extended
AA deadline tends to put other courses
immediate demands in the foreground, even if
they dont deserve to be
13
(No Transcript)
Write a Comment
User Comments (0)
About PowerShow.com