The field of Artificial Intelligence (AI) is moving towards a more transparent and explainable direction, with a focus on developing methods that can provide insights into AI decision-making processes. Recent research has emphasized the importance of explainability in AI, particularly in high-stakes applications such as healthcare and finance. One notable trend is the shift from traditional explainable AI (XAI) methods to more holistic and human-centered approaches. These new methods prioritize the needs of diverse stakeholders, including end-users, developers, and regulators, and aim to provide more comprehensive and interactive explanations. Another significant development is the increasing recognition of the need for accessibility in AI systems. Researchers are working to develop more inclusive and equitable AI systems that can be used by people with disabilities, such as visual or hearing impairments. Noteworthy papers in this area include 'Holistic Explainable AI (H-XAI): Extending Transparency Beyond Developers in AI-Driven Decision Making' which introduces a unified framework for explainability that integrates causal rating methods with traditional XAI methods, and 'From Explainable to Explanatory Artificial Intelligence: Toward a New Paradigm for Human-Centered Explanations through Generative AI' which proposes a new paradigm for explainability that leverages generative AI capabilities to provide more intuitive and interactive explanations.
Advancements in Explainable AI and Accessibility
Sources
Holistic Explainable AI (H-XAI): Extending Transparency Beyond Developers in AI-Driven Decision Making
From Explainable to Explanatory Artificial Intelligence: Toward a New Paradigm for Human-Centered Explanations through Generative AI
Accessibility Literacy: Increasing accessibility awareness among young content creators
Explainability-in-Action: Enabling Expressive Manipulation and Tacit Understanding by Bending Diffusion Models in ComfyUI
Strategies of Code-switching in Human-Machine Dialogs
Are UX evaluation methods truly accessible
Toward Machine Interpreting: Lessons from Human Interpreting Studies
EchoAid: Enhancing Livestream Shopping Accessibility for the DHH Community
Can AI Explanations Make You Change Your Mind?
StreetViewAI: Making Street View Accessible Using Context-Aware Multimodal AI
Explore, Listen, Inspect: Supporting Multimodal Interaction with 3D Surface and Point Data Visualizations
CoSight: Exploring Viewer Contributions to Online Video Accessibility Through Descriptive Commenting
Beyond Technocratic XAI: The Who, What & How in Explanation Design
RampNet: A Two-Stage Pipeline for Bootstrapping Curb Ramp Detection in Streetscape Images from Open Government Metadata
User Perception of Attention Visualizations: Effects on Interpretability Across Evidence-Based Medical Documents
Who Benefits from AI Explanations? Towards Accessible and Interpretable Systems
From Black Box to Transparency: Enhancing Automated Interpreting Assessment with Explainable AI in College Classrooms