Hervé Luga

Autonomous Virtual Actors

This paper presents the VIBES (Virtual Behaviors) framework used to simulate a ”virtual brain” capable of generating, in real time, behaviors for virtual characters. The main originality of VIBES is to combine usual behavioral animation techniques with a learning engine based on Learning Classifiers Systems in order to obtain actors that can learn how to adapt to their dynamic environment and how to efficiently combine known tasks in order to perform the user’s tasks. VIBES is a module of the V-Man [1] character animation system developed in the frame of the V-Man project supported by the European Commission in the frame of the 5th framework program.

A Synthesis of the Cell2Organ Developmental Model

Over the past two decades, many techniques have been elaborated to simulate artificial, robotic creatures at different scales. After behavioral models in the 1990s, researchers made the robot morphologies evolvable to be better adapted to their environment. More recently, developmental mechanisms of living beings have inspired “artificial embryogeny” and generated smaller creatures composed of tens to thousands of cells. Yet, there is no encompassing “transversal” model that covers multiple scales at once. To address this challenge, our project consists of growing a complete creature with various organs and high-level functionalities from a single cell. For this, we propose a developmental model, Cell2Organ, based on three simulation layers. The first layer represents a chemical solution, in which cells can divide and process substrates through chemical reactions. Its purpose is to develop a metabolism adapted to the environment and allow organisms to perform actions by using accumulated energy. We also present an alternative model for the chemical layer that replaces molecular morphogens with a generative process based on L-systems. The second layer is a hydrodynamic medium, in which cells interact with simulated substrate flows so that they can impact the whole environment. Finally, we describe our plan to extend Cell2Organ with a third, physical layer, which would allow creatures to exhibit motion in a Newtonian world. There, cells will be able to modify their individual shape and affect the overall morphology of the organism.

Cell2Organ: Self-repairing artificial creatures thanks to a healthy metabolism

For living organisms, the robustness property is capital. For almost all of them, robustness rhymes with self-repairing. Indeed, organisms are subject to various injuries brought by the environment. To maintain their integrity, organisms are able to regenerate dead parts of themselves. This mechanism, commonly named self-repairing, is interesting to reproduce. Many works exist about self-repairing in robotics and electronics but fewer are in our domain of interest, artificial embryogenesis. In this paper, we show the self-repairing abilities of our model, Cell2Organ, designed to generate artificial creatures for artificial worlds. This model has previously been presented in [1].

Gait Evolution for Humanoid Robot in a Physically Simulated Environment

This article describes a bio-inspired system and the associated series of experiments, for the evolution of walking behavior in a simulated humanoid robot. A previous study has demonstrated the potential of this approach for evolving controllers based on simulated humanoid robots with a restricted range of movements. The development of anthropomorphic bipedal locomotion is addressed by means of artificial evolution using a genetic algorithm. The proposed task is investigated using full rigid-body dynamics simulation of a bipedal robot with 15 degrees of freedom. Stable bipedal gait with a velocity of 0.94 m/s is realized. Locomotion controllers are evolved from scratch, for example neither does the evolved controller have any a priori knowledge on how to walk, nor does it have any information about the kinematics structure of the robot. Instead, locomotion control is achieved based on intensive use of sensory information. In this work, the emergence of non-trivial walking behaviors is entirely due to evolution.

ALFIL: A Crowd Simulation Serious Game for Massive Evacuation Training and Awareness

This article presents the current development of a serious game for the simulation of massive evacuations. The purpose of this project is to promote self-protection through awareness of the procedures and different possible scenarios during the evacuation of a massive event. Sophisticated behaviors require massive computational power and it has been necessary to implement several distributed programming techniques to simulate crowds of thousands of people. Even with the current state of computer hardware, the costs of building and operating this hardware is still prohibitive; so, it‘s preferred to apply distributed programming techniques running on specialized parallel computing hardware. DOI: 10.4018/ijgbl.2012070105 72 International Journal of Game-Based Learning, 2(3), 71-86, July-September 2012 Copyright

Rule fusion for the imitation of a human tutor

In virtual worlds, character credibility suffers from an increasing discrepancy between visual realism, physical modelling quality and behaviour simulation weakness. As behaviour credibility is firmly embedded in the eye of the human observer, it needs to be as close to human expectation as possible. In this study, we define a learning process able to build rule-based behaviour from the observation of a human tutor controlling a virtual agent and from a progressive fusion of rules. The ability of this imitation process to model human-controlled behaviour is assessed upon experiments carried out on a flee-attack scenario for an RTS game. Its efficiency is examined in a game development context.

Learning classifier systems and behavioural animation of virtual characters

Producing intuitive systems for the directing of virtual actors is one of the major objectives of research in virtual animation. So, it is often interesting to conceive systems that enable behavioral animation of autonomous characters, able to correctly fulfill directives from a human user considering their goal and their perception of the virtual environment. Common ways to generate behaviors of such virtual characters use usually determinist algorithm (scripts or automatons [1]). Thus the autonomy of the characters is a fixed routine that cannot adapt to novelty or any situation not previously considered. To make these virtual actors able of adaptation, we propose to combine a behavioral framework (ViBes [2]) and an evolutionist learning system, the Learning Classifier Systems [3]. Using classifiers systems we managed to make a virtual human to learn to select and to cook an aliment in order to eat something. The association of ViBes framework and two trained classifiers systems produced the following real time animation (fig. [1]) in a dynamic virtual environment.

An Introduction to the Bio-logic of Artificial Creatures

In this chapter we emphasize the most crucial developments in the design of artificial creatures starting with the seminal work of Karl Sims presented at the SIGGRAPH’94 conference and ending with the latest research in virtual beings based on cell development.

The Evolution of Artificial Neurogenesis

Evolutionary development as a strategy for the design of artificial neural networks is an enticing idea, with possible inspiration from both biology and existing indirect representations. A growing neural network can not only optimize towards a specific goal, but can also exhibit plasticity and regeneration. Furthermore, a generative system trained in the optimization of the resultant neural network in a reinforcement learning environment has the capability of on-line learning after evolution in any reward-driven environment. In this abstract, we outline the motivation for and design of a generative system for artificial neural network design.

3DEvent: a framework using event-sourcing approach for 3D web-based collaborative design in P2P

Despite recent advances, especially in web-based Collaborative Virtual Environments (CVEs) using real-time 3D content, Web technology still requires an efficient way to distribute and stream large-scale 3D data. In this paper, we present 3DEvent: an event-driven framework to collaboratively manipulate predesigned 3D content in real-time on a web-based platform. This work introduces a new approach in achieving 3D object manipulation tasks during collaborative design stages using event-sourcing. Usually, a client-server architecture supports updates to the 3D environment state. Peer-to-peer (P2P) allows direct communication between teammates reducing response times during collaboration and decreasing server load, reducing the costs of providers. 3DEvent enables P2P-assisted delivery of 3D dynamic content in a web browser via WebRTC. By combining concepts from distributed event-processing and mesh-processing, 3D independent rendering and event-based synchronization, we present 3DEvent framework and potential uses associated that support history-aware 3D applications into a unified distributed processing solution for 3D web-based CVEs.