Enhancing Predictive Analysis with Eyeknow's Time Series Models

Led the EyeKnow project, focused on detecting violent behavior between caregivers and children using computer vision and deep learning techniques. Collaborated with 50 machine learning and data professionals from around the world to develop a real-time video analysis system. Utilized Convolutional Neural Networks (CNNs) for feature extraction, Long Short-Term Memory (LSTM) networks for capturing temporal dependencies, and optical flow for motion tracking, ensuring accurate and timely violence detection in various settings, such as homes, daycare centers, and schools.