Add SeekToAction type, which will be the value for potentialAction property under VideoObject.
SeekToAction will have properties for target, which takes URL as value, and startOffset-input, which is a pseudo property a la Action schema that takes Text as value to represent timestamp value to be filled in the target.