I haven't been able to test how dialog would work with stills. If the guy in the first image would crack some joke, and it would be still image, would it work? If there were continuous background street ambience between renders, between different expressions.. I have no clue. In theory, it could work as a sort of audiobookish/cartoonish way.