Could we use this metadata to provide alternative text for the visually impaired? Or perhaps tag users on specific parts of static images?