Infant Movement

Research on infant movement focuses on automating the assessment of motor development, primarily to enable early diagnosis of neurological disorders like cerebral palsy. Current efforts leverage deep learning models, including convolutional neural networks (CNNs), recurrent neural networks (RNNs), and pose estimation algorithms (e.g., ViTPose, PoseC3D), often incorporating sensor fusion (pressure, inertial, visual) and explainable AI techniques (CAM, Grad-CAM) to improve accuracy and interpretability. These advancements hold significant promise for improving the efficiency and objectivity of clinical assessments, ultimately leading to earlier interventions and better outcomes for infants at risk.