Animation Planning for Virtual Characters Cooperation

ACM SIGGRAPH 2008 classes, 2008

2005

Elizabeth Figa of the SLIS department. I couldn't have completed this literature without her guidance. I thank Dr. Rada Mihalcea and other members of the Natural language Processing Research Group for giving me an opportunity to exhibit my work to them and discuss about it, prior to the thesis defense. My thanks also go to my friend Sudeep for helping me organize the literature. I also thank my friend Kalyan Voddi for his design contributions. The 3D models that he built served as a valuable resource for my research. I would also like to thank my friend Vivek for helping me investigate and gain knowledge about 3D animations. Last but not least, I thank all my friends who were there for me when I needed them.

2008

1994

Creating animations of a human figure performing a task requires that the agent interact with objects in the environment in a realistic way. Agent-object interaction is not completely specified from a task description alone. In this paper we sketch an architecture for the Object-Specific Reasoner (OSR), an intermediate planning module which tailors high-level plans to the specifics of the agent and objects. As plans are elaborated, the OSR generates a sequence of motion directives which are ultimately executed by a simulator. Descriptions of failures can be used to identify possible tools for the agent to use. An ObjectSpecific Reasoner is necessary in a system which allows an agent, equipped with a set of action behaviors, to interact in a semiautonomous fashion with the world. Keywords object manipulation, motion planning, animation, tools Comments Presented at Toward Physical Interaction and Manipulation, AAAI Spring Symposium Series, 1994. This conference paper is available at S...

Deformable Avatars, 2001

We present a way to mix the lower control of agents with the high level specifications of their goals. This paper addresses various topics required 10 animate virtual humans in a distributed way such as combining primary actions into tasks, using verbal communication between virtual humans and directing them with high level orders. Dur models have been tested into a multi-Ianguages / multi-modules application as described below. 1. 217 N. Magnenat-Thalmann et al. (eds.), Deformable Avatars

Engineering Letters, 2007

Realistic motion and natural-looking simulations require a thorough understanding of human movement control strategies. This work presents an initial developed on virtual agents. We describe how the agent was built searching an humanoid representation through a geometric model. Although this research is based on the work of Pettré et al. [16], we are building a solid platform to give total mobility and realism to both, the agent and the virtual environment where our humanoid works. Preliminary results about forward kinematic applied to the actor are presented. This work allow us to keep on working on virtual agents area, to analyze and propose techniques to give the agent total autonomy to its motion.

IEEE/WIC International Conference on Intelligent Agent Technology, 2003. IAT 2003., 2003

The design of a an agent system for robotics is a problem that involves aspects coming from many different disciplines (robotics, artificial intelligence, computer vision, software engineering). The most difficult part of it often consists in producing and tuning the algorithms that incorporates the robot behavior (planning, obstacle avoidance, ) and abilities (vision, manipulation, navigation, ). It is frequent that the reuse of this parts is left to a copy and paste procedure from previous applications to the new one. In so doing many problems could arises. We propose a comprehensive approach for multi-agent systems oriented to robotics applications that uses a complete design methodology supported by a specific design tools and a pattern repository that interacting each other and with the designer allow the production of a coherent design that easily incorporates patterns coming from previously experienced features and automatically produces a large part of the final code

Computer Animation and Simulation'98, 1999

This paper presents a new approach to model general interactions between virtual human agents and objects in virtual worlds. The proposed framework is designed to deal with many of the possible interactions that may arise while simulating a virtual human performing common tasks in a virtual environment. The idea is to include within the object description, all the necessary information to describe how to interact with it. For this, a feature modeling approach is used, by means of a graphical user interface program, to identify object interaction features. Moving parts, functionality instructions and interaction locations are examples of some considered features. Following this approach, the control of the simulation is decentralized from the main animation control, in the sense that some local instructions on how to deal with the object are encapsulated within the object itself. To illustrate the approach, some examples are shown and discussed.

In this paper, we present a method for an animated human agent to construct motion plans to achieve 3D-space postural goals, e.g. a goal of a hand, while avoiding collisions. We use the potential eld approach by providing mechanisms to handle the problem of local minimum. Given a conjunctive goal of multiple control points on the body, the potential eld approach tries to minimize the objective function typically dened to be a weighted sum of individual goals. The local minimum problem arises as the planner tries to locally minimize the weighted sum of individual goals, even when multiple goals of the control points do not conict with each other in 3D space. Our approach handles this problem by trying to achieve multiple goals individually, not by means of a weighted sum. To do so, the planner uses a qualitative kinematic model, which species what joint motions move what body parts in which directions in 3D space. The model is used to suggest joint motions for individual goals, and to explicitly detect and remove conicts between the suggested joint motions. The local minimum problem arises more obviously when the original goals and collision-avoidance constraints, i.e. repulsive potential elds due to obstacles conict with each other in 3D space. Our approach avoids this conict by nding intermediate postural goals of the endangered body parts, based on the kinematic simulation of the current plan.

International Journal of Social Robotics, 2010

With robotics hardware becoming more and more safe and compliant, robots are not far from entering our homes. The robot, that will share the same environment with humans, will be expected to consider the geometry of the interaction and to perform intelligent space sharing.