Interactive Decision Making for Autonomous Vehicles in Dense Traffic

Isele, David

doi:10.1109/itsc.2019.8916982

Cited by 33 publications

(20 citation statements)

References 26 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Two approaches to generate this response are used. Some studies [5,6,7,8,9,10,11,12,16,17,18] simulate human driver responses using driver models. However, many of the driver models used for this purpose are also not validated on natural human driver behavior, which could indicate a discrepancy between the simulation and natural behavior.…”

Section: Why Validate?mentioning

confidence: 99%

“…To the best of our knowledge, validation on naturalistic driving data for use in IACs has not been performed for two of the most commonly used driver models proposed for IACs. These models are the intelligent driver model IDM [23] (used in [6,7,14] to predict driver behavior and in [17,15,12] to simulate other drivers' responses) and the expected-utility-maximizing model (used e.g. in [5,4] to predict other drivers' behavior) that uses a reward function learned from human demonstrations with inverse reinforcement learning (IRL).…”

Section: Why Validate?mentioning

confidence: 99%

“…However, a major challenge for real-world implementation of AV technologies is enabling AVs to handle complex interactions with human road users. AV controllers have recently been proposed that aim to address this challenge through interaction-aware controllers (IACs) [4,5,6,7,8,9,10,11,12,13,14,15,16,17,18]. IACs incorporate a model of human driver behavior in the controller, to predict how another driver is likely to respond to the AV's behavior.…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

A human factors approach to validating driver models for interaction-aware automated vehicles

Siebinga,

Zgonnikov,

Abbink

2021

Preprint

View full text Add to dashboard Cite

A major challenge for autonomous vehicles is interacting with other traffic participants safely and smoothly. A promising approach to handle such traffic interactions is equipping autonomous vehicles with interaction-aware controllers (IACs). These controllers predict how surrounding human drivers will respond to the autonomous vehicle's actions, based on a driver model. However, the predictive validity of driver models used in IACs is rarely validated, which can limit the interactive capabilities of IACs outside the simple simulated environments in which they are demonstrated. In this paper, we argue that besides evaluating the interactive capabilities of IACs, their underlying driver models should be validated on natural human driving behavior. We propose a workflow for this validation that includes scenario-based data extraction and a two-stage (tactical/operational) evaluation procedure based on human factors literature. We demonstrate this workflow in a case study on an inverse-reinforcement-learningbased driver model replicated from an existing IAC. This model only showed the correct tactical behavior in 40% of the predictions. The model's operational behavior was inconsistent with observed human behavior. The case study illustrates that a principled evaluation workflow is useful and needed. We believe that our workflow will support the development of appropriate driver models for future automated vehicles.

show abstract

Section: Why Validate?mentioning

confidence: 99%

Section: Why Validate?mentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

A human factors approach to validating driver models for interaction-aware automated vehicles

Siebinga,

Zgonnikov,

Abbink

2021

Preprint

View full text Add to dashboard Cite

show abstract

“…The longitudinal model is an extension of the IDM model with a cooperation parameter and a perception parameter inspired by previous work [5], [6], [24]. A parameter η percept determines a yield area, if a vehicle is in this yield area, a yield action is sampled according to a Bernoulli distribution with parameter c. If c = 1, the driver always yields.…”

Section: B Level-0mentioning

confidence: 99%

“…These two methods demonstrated promising results but are unlikely to scale to very dense traffic scenarios because the search space grows exponentially with the number of agents [2]. Isele proposed a method based on game trees that requires numerous approximations in order to handle dense traffic in realtime, and the approximations can result in suboptimal decisions [5]. Recently, Bae et al suggested learning a driver model using a neural network and using that model with an online search method [6].…”

Section: Introductionmentioning

confidence: 99%

Reinforcement Learning with Iterative Reasoning for Merging in Dense Traffic

Bouton

Nakhaei²,

Isele

et al. 2020

2020 IEEE 23rd International Conference on Intelligent Transportation Systems (ITSC)

Self Cite

View full text Add to dashboard Cite

Maneuvering in dense traffic is a challenging task for autonomous vehicles because it requires reasoning about the stochastic behaviors of many other participants. In addition, the agent must achieve the maneuver within a limited time and distance. In this work, we propose a combination of reinforcement learning and game theory to learn merging behaviors. We design a training curriculum for a reinforcement learning agent using the concept of level-k behavior. This approach exposes the agent to a broad variety of behaviors during training, which promotes learning policies that are robust to model discrepancies. We show that our approach learns more efficient policies than traditional training methods.

show abstract

Quantum decision making in automatic driving

Song

Wang

et al. 2022

Sci Rep

View full text Add to dashboard Cite

The behavior intention estimation and interaction between Autonomous Vehicles (AV) and human traffic participants are the key problems in Automatic Driving System (ADS). When the classical decision theory studies implicitly assume that the behavior of human traffic participants is completely rational. However, according to the booming quantum decision theory in recent years and actual traffic cases, traffic behaviors and other human behaviors are often irrational and violate the assumptions of classical cognitive and decision theory. This paper explores the decision-making problem in the two-car game scene based on quantum decision theory and compares it with the current mainstream method of studying irrational behavior-Cumulative Prospect Theory (CPT) model. The comparative analysis proved that the Quantum Game Theory (QGT) model can explain the separation effect which the classical probability model can’t reveal, and it has more advantages than CPT model in dealing with game scene decision-making. When two cars interact with each other, the QGT model can consider the interests of both sides from the perspective of the other car. Compared with the classical probability model and CPT model, the QGT is more realistic in the behavior decision-making of ADS.

show abstract

Interactive Decision Making for Autonomous Vehicles in Dense Traffic

Cited by 33 publications

References 26 publications

A human factors approach to validating driver models for interaction-aware automated vehicles

A human factors approach to validating driver models for interaction-aware automated vehicles

Reinforcement Learning with Iterative Reasoning for Merging in Dense Traffic

Quantum decision making in automatic driving

Contact Info

Product

Resources

About