AI Supporting AAC Pictographic Symbol Adaptations

Area of Science:

• Artificial Intelligence
• Human-Computer Interaction
• Assistive Technology

Background:

• Technological advancements enable image and text manipulation for diverse applications.
• Current symbol sets for augmentative and alternative communication (AAC) lack personalization and flexibility.
• Generative AI offers potential for customized symbol creation.

Purpose of the Study:

• To explore the use of generative artificial intelligence for creating personalized pictographic symbols.
• To adapt open-licensed symbols to individual user styles and local contexts.
• To evaluate the effectiveness of AI-generated symbols through stakeholder feedback.

Main Methods:

• Utilizing image-to-image and text-to-image generative AI models.
• Adapting existing open-licensed symbol sets based on user-defined styles.
• Conducting a six-month pilot study involving symbol generation and stakeholder evaluation.
• Implementing an online voting system for outcome assessment.

Main Results:

• Successfully generated a collection of adapted symbols using AI.
• Initiated a pilot study to adapt open-licensed symbols.
• Established a framework for stakeholder evaluation of AI-generated symbols.

Conclusions:

• Generative AI holds significant potential for personalized symbol generation in AAC.
• AI-driven symbol adaptation can enhance flexibility and user experience for AAC users.
• Future work will focus on integrating successful symbols into existing AAC systems.