Simulating Heterogeneous Crowd Behaviors Using Personality Trait ...
Scalable behaviors for crowd simulation
description
Transcript of Scalable behaviors for crowd simulation
![Page 1: Scalable behaviors for crowd simulation](https://reader035.fdocuments.us/reader035/viewer/2022062520/56815f24550346895dcdf07e/html5/thumbnails/1.jpg)
SCALABLE BEHAVIORS FOR CROWD SIMULATION
By Mankyu Sung, Michael Gleicher and Stephen Chenney
![Page 2: Scalable behaviors for crowd simulation](https://reader035.fdocuments.us/reader035/viewer/2022062520/56815f24550346895dcdf07e/html5/thumbnails/2.jpg)
AUTHORS Mankyu Sung
Scalable, Controllable, Efficient and convincing crowd simulation (2005)
Michael Gleicher “I have a bad case of Academic Attention Deficit Disorder”
Stephen ChenneyFlow Tiles
![Page 3: Scalable behaviors for crowd simulation](https://reader035.fdocuments.us/reader035/viewer/2022062520/56815f24550346895dcdf07e/html5/thumbnails/3.jpg)
OUTLINE Overview Related Work Low level (probabilistic action selection) High level (situations and compositions) Results Conclusion Related Future Work Assessment
![Page 4: Scalable behaviors for crowd simulation](https://reader035.fdocuments.us/reader035/viewer/2022062520/56815f24550346895dcdf07e/html5/thumbnails/4.jpg)
OVERVIEW
Main observations: Anonymity in the
crowd what instead of who action individual
matter only in short time contribution
A character is only in a few situations at once
![Page 5: Scalable behaviors for crowd simulation](https://reader035.fdocuments.us/reader035/viewer/2022062520/56815f24550346895dcdf07e/html5/thumbnails/5.jpg)
RELATED WORK Rules based (Reynolds)
Not scalable from authoringperspective
Hierarchical (Musse)No complex individual behaviour
Physics inspired (Helbing)Limited behaviour and interaction
Annotated environment (The Sims, Kallmann)
![Page 6: Scalable behaviors for crowd simulation](https://reader035.fdocuments.us/reader035/viewer/2022062520/56815f24550346895dcdf07e/html5/thumbnails/6.jpg)
LOW LEVEL (PROBABILISTIC ACTION SELECTION)
To select new state evaluate all possible states withbehaviour function
Default behaviour functions: ImageLookup TargetFind Overlap
State:s = {t, p, θ, a, s-)
Pk(s) = 1 / (1 + e-αx)
![Page 7: Scalable behaviors for crowd simulation](https://reader035.fdocuments.us/reader035/viewer/2022062520/56815f24550346895dcdf07e/html5/thumbnails/7.jpg)
LOW LEVEL (PROBABILISTIC ACTION SELECTION)
Create complex behaviour
by composition of simple
behaviours
![Page 8: Scalable behaviors for crowd simulation](https://reader035.fdocuments.us/reader035/viewer/2022062520/56815f24550346895dcdf07e/html5/thumbnails/8.jpg)
HIGH LEVEL (SITUATIONS AND COMPOSITIONS)
Situations spatial (ATM,
crossing) non-spatial
(friendship)
When in situation: extend state graph attach sensors add event rules add behaviour
functions
Composition means union
![Page 9: Scalable behaviors for crowd simulation](https://reader035.fdocuments.us/reader035/viewer/2022062520/56815f24550346895dcdf07e/html5/thumbnails/9.jpg)
RESULTSTested on 3 scenarios: Street environment
crossing street, traffic sign, in-a-hurry Theatre environment
horizontal queue, follow, gathering, stay-in ...
Field environment follow, group, close
![Page 10: Scalable behaviors for crowd simulation](https://reader035.fdocuments.us/reader035/viewer/2022062520/56815f24550346895dcdf07e/html5/thumbnails/10.jpg)
RESULTS1,3 GHz processor 1GB
memory 500 agents with
increasing number of situations
increasing number of agents with 10 situations
![Page 11: Scalable behaviors for crowd simulation](https://reader035.fdocuments.us/reader035/viewer/2022062520/56815f24550346895dcdf07e/html5/thumbnails/11.jpg)
CONCLUSION Framework can create complex
behaviours while minimising data stored in each agent
Future work: take into account multi-agent statistics
such as crowd density more efficient simulation so not all crowd
members go through simulation step at same time
explore other mechanisms to combine behaviours to avoid time scale problem
![Page 12: Scalable behaviors for crowd simulation](https://reader035.fdocuments.us/reader035/viewer/2022062520/56815f24550346895dcdf07e/html5/thumbnails/12.jpg)
RELATED FUTURE WORK Situation Agents: Agent-based
Externalized Steering LogicSchuerman, M., Singh, S., Kapadia, M., Faloutsos P., The Journal of Computer Animation and Virtual Worlds, Special Issue CASA 2010, Wiley, pp. 1-10, 2010, in press.
Motion patches: building blocks for virtual environments annotated with motion dataLee, K. H., Choi, M. G., and Lee, J. 2006., SIGGRAPH
’06: ACM SIGGRAPH 2006 Papers, 898–906.
![Page 13: Scalable behaviors for crowd simulation](https://reader035.fdocuments.us/reader035/viewer/2022062520/56815f24550346895dcdf07e/html5/thumbnails/13.jpg)
ASSESSMENT Goals clearly specified Situation approach seems to indeed
limit the complexity of the agents Problems and possible solutions
presented Clearly structured and well written
![Page 14: Scalable behaviors for crowd simulation](https://reader035.fdocuments.us/reader035/viewer/2022062520/56815f24550346895dcdf07e/html5/thumbnails/14.jpg)
ASSESSMENTClaims and assumptions Anonymity justifies probabilistic
method?Not for low density crowds People stopping in middle of crosswalk Waiting for traffic light, then not moving
when it is green
![Page 15: Scalable behaviors for crowd simulation](https://reader035.fdocuments.us/reader035/viewer/2022062520/56815f24550346895dcdf07e/html5/thumbnails/15.jpg)
ASSESSMENTImplementation details Naive default behaviours
Path planning PRM + DijkstraPRM pre-computed, no dynamic obstacle
handlingHow are states judged to make the character
move towards position? Possible local minima? Collision detection
No prediction, possible oscillations
![Page 16: Scalable behaviors for crowd simulation](https://reader035.fdocuments.us/reader035/viewer/2022062520/56815f24550346895dcdf07e/html5/thumbnails/16.jpg)
ASSESSMENTImplementation details: extending the state graph
extension only with default graph no interaction between situations
controlling combination of behaviour functionsuse of alpha not intuitive, when to use alpha
and when to delete a behaviour
![Page 17: Scalable behaviors for crowd simulation](https://reader035.fdocuments.us/reader035/viewer/2022062520/56815f24550346895dcdf07e/html5/thumbnails/17.jpg)
ASSESSMENTLimited experiments maximum of 10 situations maximum of 500 agents random situations added, does this
include composite situations?
![Page 18: Scalable behaviors for crowd simulation](https://reader035.fdocuments.us/reader035/viewer/2022062520/56815f24550346895dcdf07e/html5/thumbnails/18.jpg)
ASSESSMENTImpact and applications Limitation on kind of applications
no evacuation simulation Situational approach might be a good
idea but should be combined with other methods
Inspiration for further research