Utilizing AI to Support in Communication with ASD Individuals
This project investigates how artificial intelligence, specifically computer vision (CV) and natural language processing (NLP), can support communication for individuals with autism spectrum disorder (ASD). The system generates simplified, meaningful narratives from user-uploaded images and adapts them to the user’s communication level using a Most-to-Least (MTL) prompting strategy drawn from applied behavior analysis (ABA). The resulting prototype, built in Streamlit and integrated with MC-LLaVA and text-to-speech (TTS), provides an accessible, low-complexity interface suitable for individuals who experience sensory or linguistic overload. Early testing demonstrated the feasibility of using a lightweight multimodal model (MC-LLaVA-3B) to produce concise captions with fewer hallucinations than larger alternatives like BLIP-2. The prototype successfully allows image upload, caption generation, audio playback, and prompt-fading controls, offering evidence that AI-assisted narrative support can be implemented in an affordable and user-adaptive way. The initiative is cost-effective due to its reliance on open-source tools and compact pre-trained models, minimizing the need for large annotated datasets or expensive infrastructure. Key lessons learned include the importance of realistic scoping (image-only input), selecting models that balance performance and computational cost, and prioritizing accessibility and user adaptability during design.
STEM with Purpose: Integrating Service Learning and Technology to Inspire Future Scientists
This presentation highlights a technology focused service-learning initiative that connects University of Houston‚ Downtown STEM undergraduates with organizations serving middle school aged youth. As Service-Learning Peer Leaders, we guide interdisciplinary teams in developing Texas Essential Knowledge and Skills -TEKS-aligned experiments that merge natural science and computer science, such as low-cost robotics builds, introductory game-design challenges, and web-based science simulations. Delivered as part of a credit-bearing course, our outreach takes place across multiple settings, including middle-school STEM events, specialized learning programs, and community organizations. This structure allows undergraduates to apply course concepts in authentic environments while supporting early STEM engagement for a wide range of learners.
Evidence of success comes from increased engagement during outreach sessions, positive feedback from teachers, and reflective student assessments showing gains in leadership, confidence, and communication skills The initiative is also intentionally cost-effective, built on inexpensive materials, open-source tools, and technologies already accessible to students, making it a scalable model even for large course sections. Core to our planning is a design matrix that maps learning objectives, materials, and costs, ensuring each experiment remains both budget-friendly and pedagogically strong. These designed experiments encourage younger students to see STEM as creative and collaborative while helping college students apply theory to community impact. We will showcase the strategies for designing cross-disciplinary, tech-infused service learning that builds curiosity, confidence, and real-world relevance in STEM education as well as our experiences.