Believable Bot Navigation via Playback of Human Traces

Abstract

Imitation is a powerful and pervasive primitive underlying examples of intelligent behavior in nature. Can we use it as a tool to help build artificial agents that behave like humans do? This question is studied in the context of the BotPrize competition, a Turing-like test where computer game bots compete by attempting to fool human judges into thinking they are just another human player. One problem faced by such bots is that of human-like navigation within the virtual world. This chapter describes the Human Trace Controller, a component of the UTˆ2 bot which took second place in the BotPrize 2010 competition. The controller uses a database of recorded human games in order to quickly retrieve and play back relevant segments of human navigation behavior. Empirical evidence suggests that the method of direct imitation allows the bot to effectively solve several navigation problems while moving in a human-like fashion.

ViZDoom: A Doom-based AI research platform for visual reinforcement learning

“…Supplying only raw visual information might relieve researchers of the burden of providing AI with high-level information and handcrafted features. We also hypothesize that it could make the agents behave more believable [16]. So far, there has been no studies on reinforcement learning from visual information obtained from FPS games.…”

MirrorBot: Using human-inspired mirroring behavior to pass a turing test

“…This category is represented by NeuroBot [7], ICE [8] (winner in 2011) and UTˆ2 [3], [9] (winner in 2012). The advantage brought by this type of approach consists in the capability of these systems to exhibit complex behavior which they previously learned from extensive human recordings.…”

UT⁁2: Human-like behavior via neuroevolution of combat behavior and replay of human traces

“…This version of the bot, referred to as UTˆ2-2010 for the rest of this paper, was based on two core ideas: (1) multiobjective neuroevolution was used to learn skilled combat behavior, but filters on the available combat actions ensured that the behavior was still human-like despite being evolved for performance, and (2) a database of traces of human play was used to help the bot get unstuck when its navigation capabilities failed. UTˆ2-2010 is described in full detail in upcoming chapters [16,20] for the book Believable Bots. The UTˆ2 bot has been modified in several ways since 2010 in order to increase its humanness rating for this year's competition: Extra input features have been provided to help it evolve better combat behavior, extra filters on combat actions inject more human knowledge into the bot, the role of human traces in the navigation of the bot has been expanded, and an extra control module has been added which encourages the bot to observe other players the way a human would, rather than simply battle them.…”