Using a Parameterized Memory Model to Modulate NPC AI

Li, Weizi Philip; Balint, Tim; Allbeck, Jan M.

doi:10.1007/978-3-642-40415-3_1

Cited by 5 publications

(2 citation statements)

References 18 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Lastly, in order to improve the safety of the AV, it is critical to model the interactions between the AV and pedestrians. There has been a number of studies making virtual humans more intelligent and life-like (Li and Allbeck, 2011;Li et al, 2012;Li and Allbeck, 2012a,b;Li et al, 2013). These results can be incorporated into the virtual environment for improving the training of the AV either directly or indirectly.…”

Section: Data Heterogeneitymentioning

confidence: 99%

Simulation and Learning for Urban Mobility: City-scale Traffic Reconstruction and Autonomous Driving

Li¹

2019

Preprint

Self Cite

View full text Add to dashboard Cite

Weizi Li: Simulation and Learning for Urban Mobility: City-scale Traffic Reconstruction and Autonomous Driving (Under the direction of Ming C. Lin) Traffic congestion has become one of the most critical issues worldwide. The costs due to traffic gridlock and jams are approximately $160 billion in the United States, more than £13 billion in the United Kingdom, and over one trillion dollars across the globe annually. As more metropolitan areas will experience increasingly severe traffic conditions, the ability to analyze, understand, and improve traffic dynamics becomes critical. This dissertation is an effort towards achieving such an ability. I propose various techniques combining simulation and machine learning to tackle the problem of traffic from two perspectives: city-scale traffic reconstruction and autonomous driving.Traffic, by its definition, appears in an aggregate form. In order to study it, we have to take a holistic approach. I address the problem of efficient and accurate estimation and reconstruction of city-scale traffic.The reconstructed traffic can be used to analyze congestion causes, identify network bottlenecks, and experiment with novel transport policies. City-scale traffic estimation and reconstruction have proven to be challenging for two particular reasons: first, traffic conditions that depend on individual drivers are intrinsically stochastic; second, the availability and quality of traffic data are limited. Traditional traffic monitoring systems that exist on highways and major roads can not produce sufficient data to recover traffic at scale. GPS data, in contrast, provide much broader coverage of a city thus are more promising sources for traffic estimation and reconstruction. However, GPS data are limited by their spatial-temporal sparsity in practice. I develop a framework to statically estimate and dynamically reconstruct traffic over a city-scale road network by addressing the limitations of GPS data.Traffic is also formed of individual vehicles propagating through space and time. If we can improve the efficiency of them, collectively, we can improve traffic dynamics as a whole. Recent advancements in automation and its implication for improving the safety and efficiency of the traffic system have prompted widespread research of autonomous driving. While exciting, autonomous driving is a complex task, consider iii the dynamics of an environment and the lack of accurate descriptions of a desired driving behavior. Learning a robust control policy for driving remains challenging as it requires an effective policy architecture, an efficient learning mechanism, and substantial training data covering a variety of scenarios, including rare cases such as accidents. I develop a framework, named ADAPS (Autonomous Driving via Principled Simulations), for producing robust control policies for autonomous driving. ADAPS consists of two simulation platforms which are used to generate and analyze simulated accidents while automatically generating labeled training data, and a hierarchical control...

show abstract

Section: Data Heterogeneitymentioning

confidence: 99%

Simulation and Learning for Urban Mobility: City-scale Traffic Reconstruction and Autonomous Driving

Li¹

2019

Preprint

Self Cite

View full text Add to dashboard Cite

show abstract

“…Memory of NPC agents has been extensively examined in the framework of decision making models. (Li, Balint, and Allbeck 2013) examines the use of a parameterized model of knowledge, which allows for memory and forgetting, in order to allow for human like behaviours based on memories in NPC agents. Mooney et al (Mooney and Allbeck 2014) look at NPC reputations of players based on a Bayesian model; previous interactions with the player are monitored and expected future actions are predicted.…”

Section: Introductionmentioning

confidence: 99%

Reputation Systems for Non-Player Character Interactions Based on Player Actions

Brown

Lee

Kraev

2021

AIIDE

View full text Add to dashboard Cite

In digital games there is an emphasis on the idea of quest completion; by completing a quest the character wins resources from the giver of the quest, they also will gain a reputation among the Non-Player Characters (NPCs) for its completion. However, this reputation currently propagates across the game world in an unrealistic manner; many NPCs will know of a completion of a quest many townships over without a narrative rationale. In this paper, we examine a method for allowing NPC interactions to spread reputation in a game world from an initial witness point of a quest completion to all other NPCs. This model is examined in a series of connected graphs: size five models, small world graphs, and graphs developed from digital games. Tests show that propagation of the information is highly dependent upon easily established properties of interactions, such as the graph regularity, average degree, and diameter. Further, real game graphs demonstrate that information generated in high population hubs can be propagated faster than that generated in smaller quests from outlying areas.

show abstract

Functional Crowds

Allbeck¹

2018

Handbook of Human Motion

View full text Add to dashboard Cite

Using a Parameterized Memory Model to Modulate NPC AI

Cited by 5 publications

References 18 publications

Simulation and Learning for Urban Mobility: City-scale Traffic Reconstruction and Autonomous Driving

Simulation and Learning for Urban Mobility: City-scale Traffic Reconstruction and Autonomous Driving

Reputation Systems for Non-Player Character Interactions Based on Player Actions

Functional Crowds

Contact Info

Product

Resources

About