You need to process and analyze large volumes of video data in real-time to detect and identify objects, people, text, and activities. Which AWS service would you use?