We propose a data-driven approach to automatically generate a scene where tens to hundreds of characters densely interact with each other. During off-line processing, the close interactions between characters are precomputed by expanding a game tree, and these are stored as data structures called interaction patches. Then, during run-time, the system spatio-temporally concatenates the interaction patches to create scenes where a large number of characters closely interact with one another. Using our method, it is possible to automatically or interactively produce animations of crowds interacting with each other in a stylized way. The method can be used for a variety of applications including TV programs, advertisements and movies.

No abstract

The concept of response threshold (RT) has been developed to explain task allocation in social insect colonies, wherein individual workers engage in tasks depending on their responsiveness to the task-related stimulus. Moreover, a mathematical model of RT has been proposed to explain data obtained from task allocation experiments; however, its applicability range warrants clarification through adequate quantitative analysis. Hence, we used an automatic measuring system to count passage events between a nest chamber and a foraging arena in five colonies of ants, Camponotus japonicus . The events were measured using radio-frequency identification tags attached to all workers of each colony. Here, we examined the detailed forms of i) labour distribution during foraging among workers in each colony and ii) the persistence of rank-order of foraging among workers. We found that labour distribution was characterized by a generalized gamma-distribution, indicating that only few workers carried out a large part of the workload. The rank-order of foraging activity among workers in each colony was maintained for a month and collapsed within a few months. We compared the obtained data with testable predictions of the RT model. The comparison indicated that proper evaluation of the mathematical model is required based on the obtained data.

a) (b) (c) (d) Figure 1: Multi-character animations are synthesized from single-person Motion Capture data. The individual interactions between nearby characters are precomputed into interaction patches by expanding game trees during the off-line processing stage. Our system automatically concatenates the patches and generates a complex multi-character animation, such as (a) one person fighting with many enemies, (b) a group of characters falling down onto each other like dominos, (c) an American football player holding a ball and escaping from tackling defenders, and (d) a group of people passing luggage from one to another. AbstractWe propose a data-driven approach to automatically generate a scene where tens to hundreds of characters densely interact with each other. During off-line processing, the close interactions between characters are precomputed by expanding a game tree, and these are stored as data structures called interaction patches. Then, during run-time, the system spatio-temporally concatenates the interaction patches to create scenes where a large number of characters closely interact with one another. Using our method, it is possible to automatically or interactively produce animations of crowds interacting with each other in a stylized way. The method can be used for a variety of applications including TV programs, advertisements and movies.

No abstract