Face Touch

Research on face touching focuses on automatically detecting this behavior using computer vision and wearable sensor technologies, primarily to understand its role in disease transmission and neurodevelopment. Current approaches leverage deep learning models, including supervised contrastive learning and random forests, analyzing video data (body posture, hand movements) or accelerometer/gyroscope data from smartwatches to achieve high accuracy in face touch detection. This work has implications for public health initiatives (e.g., infectious disease control) and early diagnosis of neurodevelopmental disorders by providing objective, quantifiable measures of infant behavior.