Google DeepMind Unveils V2A AI for Generating Unlimited Video Soundtracks and Dialogue

Jun 18, 2024, 02:02 AM

Google DeepMind has unveiled its latest advancement in AI technology with the introduction of video-to-audio (V2A) generative technology. This new AI tool can generate an unlimited number of synchronized soundtracks and dialogue for silent videos using video pixels and text prompts. The technology aims to add sound that matches the acoustics of the scene and accompanies on-screen action, potentially revolutionizing video creation workflows. While the results are not yet fully convincing, the technology showcases significant progress in AI's ability to create audio content. Google DeepMind has shared multiple examples, including 13 examples, demonstrating the capabilities of this technology, although there are no plans for public release at this time. The videos presented were initially generated with Google's video generation model, Veo.