Fetching the paper…

Connecting the Dots between Audio and Text without Parallel Data through Visual Knowledge Transfer · Around