Physical Based AnimationSimulation

1
Physical Based Animation/Simulation
2
Particle Systems

• Particle systems offer a solution to modeling
  amorphous, dynamic and fluid objects like clouds,
  smoke, water, explosions and fire.

3
Representing Objects with Particles

• An object is represented as clouds of primitive
  particles that define its volume rather than by
  polygons or patches that define its boundary.
• A particle system is dynamic, particles changing
  form and moving with the passage of time.
• Object is not deterministic, its shape and form
  are not completely specified. Instead

4
Basic Model of Particle Systems

• New particles are generated into the system.
• Each new particle is assigned its individual
  attributes.
• Any particles that have existed past their
  prescribed lifetime are extinguished.
• The remaining particles are moved and transformed
  according to their dynamic attributes.
• An image of the particles is rendered in the
  frame buffer, often using special purpose
  algorithms.

5
Particle Attributes

• Initial position
• Initial velocity
• Initial size
• InitialSize MeanSize Rand() X VarSize
• Initial color
• Initial transparency
• Shape
• Lifetime

AliasWavefronts Maya
6
Particle Dynamics

• A particles position is found by simply adding
  its velocity vector to its position vector. This
  can be modified by forces such as gravity.

• Other attributes can vary over time as well, such
  as color, transparency and size. These rates of
  change can be global or they can be stochastic
  for each particle.

7
Particle Extinction

• When generated, given a lifetime in frames.
• Lifetime decremented each frame, particle is
  killed when it reaches zero.
• Kill particles that no longer contribute to image
  (transparency below a certain threshold, etc.).

8
Particle Rendering

• Particles can obscure other objects behind them,
  can be transparent, and can cast shadows on other
  objects. The objects may be polygons, curved
  surfaces, or other particles.

9
Star Trek II The Wrath of Khan
10
Particle Hierarchy

• Particle system such that particles can
  themselves be particle systems.
• The child particle systems can inherit the
  properties of the parents.

11
Grass

• Entire trajectory of a particle over its lifespan
  is rendered to produce a static image.
• Green and dark green colors assigned to the
  particles which are shaded on the basis of the
  scenes light sources.
• Each particle becomes a blade of grass.

white.sand by Alvy Ray Smith (he was also working
at Lucasfilm)
12
Soft Bodies

• Particle system deforms the surface of a NURBS or
  polygonal object.

chewing gum soft body
13
Physical Based Animation/Simulation
14
(No Transcript)
15
Flocking

• Schooling or swarming or herding
• Relate to groups of characters
• Craig W. Reynolds, Flocks, herds and schools A
  distributed behavioral model, SIGGRAPH 87
• Three simple rules (steering behavior)
• Separation, Alignment, Cohesion
• Together gives groups of autonomous agents
  (boids) a realistic form of group behavior
  similar to flocks of birds, schools of fish, or
  swarms of bees. ex1, ex2
• The steering behavior determines how a character
  reacts to other characters in its local
  neighborhood.

Birds plus -oids
16
Emergent Behaviors

• Combination of three flocking rules results in
  emergence of fluid group movements
• Emergent behavior
• Behaviors that arent explicitly programmed into
  individual agent rules
• Ants, bees, schooling fishes

17
Three Rules (Steering Behaviors)

• Separation steer to avoid crowding local
  flockmates
• Alignment steer toward the average heading of
  local flockmates
• Cohesion steer to move toward the average
  position of local flockmates

18
Three Rules (Steering Behaviors)

• In each rule, the steering behavior determines
  how a character reacts to other characters in its
  local neighborhood.
• Characters outside of the local neighborhood are
  ignored.
• The neighborhood is specified by a distance which
  defines when two characters are nearby, and an
  angle which defines the characters perceptual
  field of view.

19
Separationsteer to avoid crowding local
flockmates
Gives a character the ability to maintain a
certain separation distance from others nearby.
20
How to Compute Steering for Separation?

• First a search is made to find other characters
  within the specified neighborhood (exhaustive,
  spatial partitioning, caching scheme)
• For each nearby character, a repulsive force is
  computed by subtracting the positions of our
  character and the nearby character, normalizing,
  and then applying a 1/r weighting. (That is, the
  position offset vector is scaled by 1/r 2.)
• These repulsive forces for each nearby character
  are summed together to produce the overall
  steering force.

21
Alignmentsteer toward the average heading of
local flockmates
Gives an character the ability to align itself
with (that is, head in the same direction and/or
speed as) other nearby characters
22
How to Compute Steering for Alignment?

• Find all characters in the local neighborhood (as
  described for separation)
• Average together the velocity (or alternately,
  the unit forward vector) of the nearby
  characters.
• This average is the desired velocity, and so
  the steering vector is the difference between the
  average and our characters current velocity (or
  alternately, its unit forward vector).
• This steering will tend to turn our character so
  it is aligned with its neighbors.

23
Cohesionsteer to move toward the average
position of local flockmates
Gives an character the ability to cohere with
(approach and form a group with) other nearby
characters
24
How to Compute Steering for Cohesion?

• Find all characters in the local neighborhood (as
  described for separation)
• Computing the average position (or center of
  gravity) of the nearby characters.
• The steering force can applied in the direction
  of that average position (subtracting our
  character position from the average position, as
  in the original boids model), or it can be used
  as the target for seek steering behavior.

25
Separation, Alignment and Cohesion

• In some applications it is sufficient to simply
  sum up the three steering force vectors to
  produce a single combined steering for flocking
• However for better control it is helpful to
• normalize the three steering components
• scale them by three weighting factors before
  summing them.
• As a result, boid flocking behavior is specified
  by nine numerical parameters
• a weight (for combining),
• a distance and an angle (to define the
  neighborhood)
• for each of the three component behaviors.

26
Combined Behaviors and Groups

• Flocking (combining separation, alignment,
  cohesion)
• Crowd Path Following
• Leader Following
• Unaligned Collision Avoidance
• Queuing (at a doorway)

27
Physical Based Animation/Simulation
28
Cognitive Modeling
Use AI to allow for planning and learning
29
Control Algorithms
Simplified control loop

• Use feedback to maintain
• balance
• velocity (speed and direction)
• etc.

30
State Machines
Separate the motion or behavior into several
simple states
Simple states allow us to generate laws
State transitions are triggered by events
Example fall forward until foot hits the ground
31
Running State Machine
32
Overview

• Virtual Creatures
• Creature Representation
• Creature Control
• Physical Simulation
• Behavior
• Evolution
• Results

33
Virtual Creatures

• Complexity vs. Control
• Genetic Algorithms
• Darwin (fitness)
• Differs from previous work

34
Creature Representation

• Phenotype

• Genotype

35
Creature Representation

• Directed Graph
• Nodes
• Information
• Dimensions
• Joint-type
• Joint-limits
• Recursive-limit
• Neurons
• Connections
• Child Node
• Position
• Orientation

36
Creature Control

• Brain
• A directed graph of neurons
• Effectors
• Applied at Joints as Forces or Torques
• Muscle Pairs

37
Creature Control

• Neurons
• Provide different functions
• Sum, product, abs, max, sin, cos, oscillators,
  etc
• Output vs. Input
• Number of inputs dependant on function
• Output dependant on input and maybe previous state

38
Combining Control and Representation
39
Physical Simulation

• Collision Detection
• Bounding Box Pair Specific
• Collision Response
• Impulses penalty springs
• Friction
• Viscosity
• For simulating underwater

40
Behavior

• Evolution for a specific behavior
• Swimming
• Walking
• Jumping
• Following (Land/Water)
• Fitness function evaluated at each step
• Weights for more preferred methods

41
Evolution

• Recipe for a successful evolution
• Create initial genotypes
• From scratch
• Calculate survival ratio
• Evaluate fitness and kill off the weaklings
• Reproduce the most fit
• Evolve, and proceed to step 3.

42
Evolution

• Mating CrossOver Mutation

• Reproductive Method
• 40 Asexual
• 30 Crossover
• 30 Grafting

43
Performance

• CM-5 with 32 processors 3 Hours
• Population of 300
• 100 Generations

44
Results

• Homogeneity
• Swimmers
• Paddlers
• Tail-waggers
• Walkers
• Lizard-like
• Pushers/Pullers
• Hoppers
• Followers
• Steering Fins
• Paddlers

45
Overview of vBeluga

• Virtual belugas are shown in a wild pod context
• Incorporates research on beluga behavior and
  vocalization conducted at aquarium UBC Zoology
• Flow scientist game visitors wild belugas
  captive - wild
• Simulation AI architecture - belugas can learn
  and alter their behavior based on changes in
  their environment updatable new scientific
  thinking
• Physically-based system allows for natural whale
  locomotion and realistic water game research
• Realistic graphics use of actuators (virtual
  bones and muscles) - game research

46
Beluga Behavior SystemNNet, Action Selection
DiPaola,Akai,Kraus 06 "Experiencing Belugas
Developing an Action Selection-Based Aquarium
Interactive", Journal of Adaptive Behavior
Foundation AI (NSERC) DiPaola,Akai 06 Designing
Adaptive Multimedia Interactives to Support
Shared Learning Experiences", ACM Siggraph
Education Design HCI / Informal Learning
(SAGE) DiPaola,Akai 06 "Blending Science
Knowledge and AI Gaming Techniques for
Experiential Learning", CA Game Studies Assoc.
Gaming/Learning DiPaola,Akai 05 Shifting
Boundaries the Ontological Implications of
Simulating Marine Mammals, NewForms, Museum of
Anthropology IT/Society
47
Vancouver Aquarium Adv. Layer
48
Neural Net Layer
49
Flexibility of use

• Decouple Display with UI (tabletop)
• Control crowd by related placement
• Main gallery full ui tabletop,
  projection, signage
• Summer camp simple ui on
  every system
• Beluga encounters guided system
• Corporate gathering main system, ambient
  mode

50
(No Transcript)
51
Physical Based Animation/Simulation