Posture Boost: Promoting children’s health through real-time, AI-based posture analysis from an early age
Can camera-based AI detect and classify sitting postures? Which AI models can help determine poor posture? Can these models be applied in real teaching situations?
About the Project
Imagine a classroom where every child learns not just to read and write, but also to sit and grow healthy—with a little help from AI. An ergonomically correct body posture in children is crucial to prevent long-term health issues related to posture and support effective learning. The aim of Posture Boost is to address the lack of concepts for learning healthy posture from an early age.
The project focuses on developing an AI-based app with real-time posture analysis to promote healthy sitting posture while writing and to foster body awareness. This approach involves children, parents, and teachers to enable them to take the necessary measures to maintain healthy posture at home and at school as early as possible. The app’s posture analysis uses the smartphone or tablet camera as a sensor and a machine learning model that detects and categorizes the posture based on the photos taken. Subjective assessments of therapists are used to create the AI model, which is intended to reproduce the subjective assessment of therapists. The results should demonstrate the reliable and valid use of the AI-based app for analyzing posture under everyday conditions.
Posture Boost will help children develop a healthy awareness of their own posture. By helping children discover their own posture today, we are helping to shape a healthier generation for tomorrow.
About the Fellow
Kathrin Schmalzl is a PhD candidate at the TUM School of Medicine and Health. She focuses on children’s postural behavior and how AI- based approaches can improve posture and body awareness, with an emphasis on practical application in everyday life.
