Owli-AI Research · Open Source · Accessibility
Owli-AI is a non-commercial research and development initiative from the environment of Technische Hochschule Mittelhessen.
A volunteer team develops and evaluates AI-supported accessible technologies for education and everyday life – from scientific methods to open-source prototypes and assistive apps.
Updates
Project launch
The publicly funded THM research and transfer project develops AI-supported methods for accessible STEM teaching materials. Detailed project communication is maintained on the dedicated IncluLearn.AI website.
Award
Justine Buss's bachelor's thesis was awarded the THM Karl-Heinz Lust Innovation Prize on 1 October 2026. The work also led to a workshop contribution presented at ECCV 2026 in Malmö.
Research & transfer
Owli-AI connects research on accessible STEM materials and assistive computer vision with open-source prototypes and applications. New results are linked to their underlying projects and publications.
New in Owli-AI Assist
Capture or choose a picture, let Owli-AI Assist create a 30-second English song, and share the result as a video.
Beta feature: AI results can vary.
Available now
The apps can now be tried for free on Google Play.
Your visual assistance system with AI support.
Analyze scenes, read text, ask follow-up questions, and hear answers with profiles for overview, detail, and OCR.
The magnifier for everything that is too small in daily life.
Magnification, contrast filters, and read-aloud features, your everyday magnifier on Android.
Owli-AI is designed for:
The applications are modular and can be used individually or in combination.
Each app fulfills a clearly defined task. Together they form a continuous assistance system.
Digital orientation aid for detecting walkable paths with acoustic feedback.
AI-supported scene analysis with follow-up questions, voice control, and context-based assistance.
Mobile magnifier with contrast filters, OCR, and magnifying mirror. Now available for free on Google Play.
Real-time transcription and optimized speech output for better speech intelligibility.
Desktop magnification with integrated reading features, currently for Windows, Linux in planning.
Owli-AI is continuously developed. The goal is to connect mobile and desktop assistance even more closely and combine intelligent functions in a meaningful way.
New
Short audio summaries from NotebookLM and additional songs offer a compact introduction to Owli-AI projects, background topics, and research.
In addition to the app suite, specialized custom GPTs are available in ChatGPT.
Social law in Hesse made understandable: responsibilities, deadlines, notices, plus support for drafting letters.
Structured image description, orientation, and OCR in ChatGPT for blind and severely visually impaired people.
Accessible CE-1 tutor: step by step, screen-reader friendly, with linear descriptions instead of left/right references.
Detects walkable paths and gives acoustic direction cues directly on your smartphone.
Analyze scenes, read text, ask follow-up questions, and hear answers with profiles for overview, detail, and OCR.
Magnification, contrast filters, and read-aloud features, your everyday magnifier on Android.
Real-time transcription, simplified language, and clear TTS output for better intelligibility.
Desktop magnification with contrast and color modes for Windows, Linux in planning.