Generating Believable Virtual Characters Using Behavior Capture and Hidden Markov Models

  title={Generating Believable Virtual Characters Using Behavior Capture and Hidden Markov Models},
  author={Richard Zhao and Duane Szafron},
  booktitle={Advances in Computer Games},
We propose a method of generating natural-looking behaviors for virtual characters using a data-driven method called behavior capture. We describe the techniques for capturing trainer-generated traces, for generalizing these traces, and for using the traces to generate behaviors during game-play. Hidden Markov Models (HMMs) are used as one of the generalization techniques for behavior generation. We compared our proposed method to other existing methods by creating a scene with a set of six… 

Virtual character behavior architecture using cyclic scheduling

This research proposes a new Tiered Behavior Architecture model for controlling the behaviors of virtual characters and describes experiments and user studies that validate this model.

Evolving Artificial Neural Networks applied to generate virtual characters

ADANN is presented, a system for the automatic evolution and adaptation of artificial neural networks based on evolutionary ANN (EANN), which use Genetic Algorithm (GA) that evolves fully connected Artificial Neural Network.

Exploring Apprenticeship Learning for Player Modelling in Interactive Narratives

This paper used Receding Horizon IRL to learn players' goals in the form of reward functions, and derive policies to imitate their behaviour, and suggested that RHIRL is able to learn action sequences to complete a game, and provided insights towards generating behaviour more similar to specific players.

Effects of Gender on Perception and Interpretation of Video Game Character Behavior and Emotion

It is found that female gamers were more appreciative of complex behaviors than male gamers, and if game designers want players to perceive these characters in a certain way, they should consider the gender of targeted players.

Writing Real-Time .Net Games in Casanova

In this paper we show the Casanova language (and its accompanying design pattern, Rule-Script-Draw) in action by building a series of games with it. In particular we discuss how Casanova is suitable



Using behaviour patterns to generate scripts for computer role-playing games

This research describes the design and implementation of a novel behaviour model for interacting NPCs, based on generative design patterns, that requires no manual script writing and produces scripting code for BioWare Corp.'s Never-winter Nights computer role-playing game.

Automatic learning and generation of social behavior from collective human gameplay

Promising results are documents from the first iteration of a Collective Artificial Intelligence system that generates behavior and dialogue in real-time from data captured from over 11,000 players of The Restaurant Game.

Animation from observation: Motion capture and motion editing

Animation is a uniquely expressive art form: it provides the creator with control over both the appearance and the movement of characters and objects, which when well used, can create works with tremendous impact.

Predicting player behavior in Tomb Raider: Underworld

Results reveal that linear regression models and other non-linear classification techniques perform well on the tasks and that decision tree learning induces small yet well-performing and informative trees.

Learning from Human Demonstrations for Real-Time Case-Based Planning

This paper presents a collection of algorithms that can be used to automatically learn plans from human demonstrations based on the basic idea of a plan dependency graph, which is a graph that captures the dependencies among actions in a plan.

An Architecture for Game Behavior AI: Behavior Multi-Queues

An AI behavior architecture that supports responsive collaborative interruptible and resumable behaviors using behavior queues using a simple efficient mechanism for encapsulating behaviors into components that can change dynamically, based on environmental criteria is described.

Imitation learning at all levels of game-AI

If and how the principle of imitation learning can facilitate the programing of life-like computer game charecters and different algorithms that learn from human generated training data are presented and it is shown that machine learning can be applied on different levels of cognitive abstraction.