From 10 to 6 minutes per session
Accurate emotional interpretation
Patient-therapist
Across 5 implemented clinics
Innovative AI system that analyzes behavioral patterns to facilitate non-verbal communication in autistic children
The system uses open-source NLP models and visual recognition to interpret gestures, expressions and vocalizations. Reduces signal decoding time by 40% and improves interactions by 25%.
Grok, LLaMA and Qwen models for interpreting vocalizations and linguistic patterns with 82% accuracy in behavioral analysis.
YOLOv5 for facial expression and gesture analysis, reducing signal decoding time from 10 to 6 minutes per session.
React Native mobile app for therapists and families supporting 1,000 simultaneous users for monitoring and interventions.
Real-world applications of the AI system for autism in clinical and educational contexts
Support for therapists in interpreting non-verbal signals during sessions, improving therapeutic intervention effectiveness.
Integration in school environments to facilitate communication between autistic students and teachers, promoting inclusion.
Mobile app for parents and caregivers providing insights into behavioral patterns and personalized communication improvement suggestions.
Grok, LLaMA, Qwen for behavioral analysis
YOLOv5 for visual recognition
Mobile app for therapists and families
Scales for 1,000 simultaneous users
A pediatric clinic needed a system to improve communication with autistic children, as described in the presentation.
X Quantum Tech developed an AI system that analyzes behavioral patterns (gestures, expressions, vocalizations) using open-source NLP models and visual recognition, with mobile app for therapists and families.