ViDA-MAN: Visual Dialog with Digital Humans
This addresses the need for more human-like and immersive interactions in digital assistants, though it appears incremental by integrating existing multi-modal techniques.
The paper tackles the problem of creating a digital-human agent for multi-modal interaction, resulting in ViDA-MAN, which offers real-time audio-visual responses to speech inquiries with sub-second latency and high-quality videos.
We demonstrate ViDA-MAN, a digital-human agent for multi-modal interaction, which offers realtime audio-visual responses to instant speech inquiries. Compared to traditional text or voice-based system, ViDA-MAN offers human-like interactions (e.g, vivid voice, natural facial expression and body gestures). Given a speech request, the demonstration is able to response with high quality videos in sub-second latency. To deliver immersive user experience, ViDA-MAN seamlessly integrates multi-modal techniques including Acoustic Speech Recognition (ASR), multi-turn dialog, Text To Speech (TTS), talking heads video generation. Backed with large knowledge base, ViDA-MAN is able to chat with users on a number of topics including chit-chat, weather, device control, News recommendations, booking hotels, as well as answering questions via structured knowledge.