Artificial Contender™: bringing more life into gameplay with behavior-capture AI

Today's games are no longer about just the latest 3D graphics. As visual quality is getting closer to photo realism, customer expectations of gameplay level grow proportionately. Now end-users will often judge a game by the sophistication of interaction between characters and expect game agents to be as 'human' as possible in terms of behavior. Ideally, gamers want to see a game character learn, acquire a specific style, even make mistakes during the game the way real humans do.

Though the quality and depth of AI have improved tremendously in recent years, there remains a strong need for better AI and new AI-related gaming features. The new generation of hardware now has the necessary resources for advanced AI methods allowing these features. But until recently an integrated solution enabling game developers to achieve a desirable level of AI sophistication, while meeting their budgets and timeframe limits, remained elusive.

One of promising emerging trends in AI methods is behavior-capture AI – an AI that can learn a human player's behavior and style. The term is derived from ‘motion capture’ because of the similar concept – real humans are used as models for game agents.

Realizing the opportunity to create an AI solution that would take gameplay to new heights of sophistication and excitement, TruSoft International Inc. undertook the development of Artificial Contender™, a groundbreaking behavior-capture AI technology and AI middleware allowing new gaming modes, more dynamic game play, and new tools for game developers.

Showcase: Artificial Contender™ in the development of SONY's This Is Football
AC Demos