Demonstrating the use of Machine Learning to enable creation of "virtual" cameras from a single high-resolution camera input. A single high resolution 4K camera input is analyzed to identify faces and automatically track, crop and create multiple High Definition (HD) virtual streams.