Our Approach
We collect speech samples from volunteers who stutter, ensuring data privacy and ethical compliance. Each sample is analyzed and annotated by certified speech-language pathologists. Our methods adhere to international research ethics standards with approvals obtained from the participants.
Key Objectives
- Develop an AI-based system to automatically detect and assess stuttering severity.
- Build and expand an annotated database of speech samples from people who stutter.
- Validate AI models through clinical collaborations.
Impact
- Provides real-time stuttering severity analysis.
- Enhances speech therapy treatment precision.
- Contributes to global research on communication disorders.
- Facilitates clinical trials for pharmaceutical research.
Methodology
📊 Data Collection
- Audiovisual samples of Arabic & English speech from stuttering individuals.
- Manual annotation by speech-language pathologists.
🤖 Model Development
- Initial AI models trained on annotated data.
- Continuous improvement through machine learning iterations.
✅ Validation & Impact Assessment
- Comparing AI-generated assessments with human expert ratings.