Explainable artificial intelligence for autonomous driving: An overview and guide for future research directions

Atakishiyev, Shahin; Salameh, Mohammad; Yao, Hengshuai; Goebel, Randy

doi:10.48550/arxiv.2112.11561

Cited by 28 publications

(35 citation statements)

References 0 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…However, we need to test our system on many more interesting scenarios so that we can generate a more varied set of explanations if we want to be certain that XAVI does indeed work properly and is useful for passengers. Besides the scenarios by Albrecht et al (2021), we can base further evaluation on the scenarios presented by Wiegand et al (2020) which were specifically collected to evaluate XAI in autonomous driving scenarios. In the future, it will also be important to run a user study on how the generated explanations affect trust and knowledge levels in humans, as our ultimate goal is to achieve trustworthy autonomous driving.…”

Section: Discussionmentioning

confidence: 99%

“…We propose a human-centric explanation generation method called eXplainable Autonomous Vehicle Intelligence (XAVI), focusing on high-level motion planning and prediction. XAVI is based on an existing transparent, inherently interpretable motion planning and prediction system called IGP2 [Albrecht et al, 2021]. An example of the output of our system is shown in Figure 1.…”

Section: Introductionmentioning

confidence: 99%

“…However, instance-based explanations may ignore the overall workings of the black box and, when interpreted incorrectly, may build an incorrect understanding of our system in humans. Global surrogate models instead use black box models as a teacher to train simpler, interpretable white box systems [Bastani et al, 2017;Lakkaraju et al, 2017]. These can model the overall decision process of the teacher, though usually at the cost of soundness.…”

Section: Introductionmentioning

confidence: 99%

“…Recently, Albrecht et al (2021) proposed an inherently interpretable integrated planning and prediction system called 1 https://github.com/uoe-agents/xavi-ai4ad IGP2. This method relies on intuitive high-level driving actions and uses rational inverse planning to predict the trajectories of other vehicles, which are then used to inform motion planning with Monte Carlo Tree Search (MCTS).…”

Section: Introductionmentioning

confidence: 99%

See 3 more Smart Citations

A Human-Centric Method for Generating Causal Explanations in Natural Language for Autonomous Vehicle Motion Planning

Gyevnar¹,

Tamborski²,

Wang³

et al. 2022

Preprint

View full text Add to dashboard Cite

Inscrutable AI systems are difficult to trust, especially if they operate in safety-critical settings like autonomous driving. Therefore, there is a need to build transparent and queryable systems to increase trust levels. We propose a transparent, human-centric explanation generation method for autonomous vehicle motion planning and prediction based on an existing white-box system called IGP2. Our method integrates Bayesian networks with context-free generative rules and can give causal natural language explanations for the highlevel driving behaviour of autonomous vehicles. Preliminary testing on simulated scenarios shows that our method captures the causes behind the actions of autonomous vehicles and generates intelligible explanations with varying complexity.

show abstract

Section: Discussionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

A Human-Centric Method for Generating Causal Explanations in Natural Language for Autonomous Vehicle Motion Planning

Gyevnar¹,

Tamborski²,

Wang³

et al. 2022

Preprint

View full text Add to dashboard Cite

show abstract

“…Longo et al [60] and Ahmed et al [61] also reviewed XAI applied in industry. In the automotive field, explainable AI is used to create transparency and understand model decisions [62], but it turns out that XAI itself is not sufficient to increase the trust [63]. Other explanations need to be provided, not only for developers, but also for the end-user.…”

Section: Review Of Explainable Industrial Ai Applicationsmentioning

confidence: 99%

Privacy-Preserving and Explainable AI in Industrial Applications

et al. 2022

View full text Add to dashboard Cite

The industrial environment has gone through the fourth revolution, also called “Industry 4.0”, where the main aspect is digitalization. Each device employed in an industrial process is connected to a network called the industrial Internet of things (IIOT). With IIOT manufacturers being capable of tracking every device, it has become easier to prevent or quickly solve failures. Specifically, the large amount of available data has allowed the use of artificial intelligence (AI) algorithms to improve industrial applications in many ways (e.g., failure detection, process optimization, and abnormality detection). Although data are abundant, their access has raised problems due to privacy concerns of manufacturers. Censoring sensitive information is not a desired approach because it negatively impacts the AI performance. To increase trust, there is also the need to understand how AI algorithms make choices, i.e., to no longer regard them as black boxes. This paper focuses on recent advancements related to the challenges mentioned above, discusses the industrial impact of proposed solutions, and identifies challenges for future research. It also presents examples related to privacy-preserving and explainable AI solutions, and comments on the interaction between the identified challenges in the conclusions.

show abstract

Selecting Explanation Methods for Intelligent IoT Systems: A Case-Based Reasoning Approach

Parejas-Llanovarced

Darias

Caro-Martínez

et al. 2023

Lecture Notes in Computer Science

View full text Add to dashboard Cite

Explainable artificial intelligence for autonomous driving: An overview and guide for future research directions

Cited by 28 publications

References 0 publications

A Human-Centric Method for Generating Causal Explanations in Natural Language for Autonomous Vehicle Motion Planning

A Human-Centric Method for Generating Causal Explanations in Natural Language for Autonomous Vehicle Motion Planning

Privacy-Preserving and Explainable AI in Industrial Applications

Selecting Explanation Methods for Intelligent IoT Systems: A Case-Based Reasoning Approach

Contact Info

Product

Resources

About