Creating captions through speech-to-text