Novel approach for learning visual representations using caption annotations for improved visual understanding.
Jan 1, 2020