Transcribe speech from audio or video files, automatically extracting audio and converting to text using Docker Whisper ASR for Douyin/TikTok media.