• exploiting the Compute Unified Device Architecture (CUDA) to improve real-time object segmentation and tracking performance and to reduce error rates,
• using CUDA for machine learning and training on massive amounts of data to improve classification and identification of user gestures and postures, identifying and segmenting various users in the field of view,
• learning users' behavior and habits,
• differentiating users through various biometrics, and
• performing usability testing on thousands of subjects.