Encoding and Controlling Global Semantics for Long-form Video Question Answering

Open in new window