Physical Address
304 North Cardinal St.
Dorchester Center, MA 02124
The last AI -born AI, VEO 3, can create a sound to keep up with the clips that create.
Tuesday During the Google I/O 2025 developer conferenceGoogle unveiled the VEO 3, which the company claims to generate sound effects, background noise, and even a dialogue to accompany the videos that create it. VEO 3 also improved its predecessor, I see 2In terms of the quality of the shots you can create, Google says.
VEO 3 on Tuesday is available in the Google Chatbot app for subscribers Google Plan at $ 249.99 a monthWhere it can be demanded for text or image.
“For the first time, we get out of the silent video generation.” ((You can give VEO 3) a claim describing the letters and the environment, and a dialogue suggestion with a description of how they want to look. “
The availability of widespread tools for the creation of video generators led to an explosion of service providers so that the area becomes saturated. Startups Runningand LightricksGenmo, Beccaand HigsfieldAnd kling and LumaAs well as technology giants such as Openai And albaba, make models in a quick clip. In many cases, a little is distinguished from one model.
The sound output will be a great discrimination for the Veo 3, if Google can make its promises. Acting sound generation tools no a novelAnd no models to create video voice Antiquities. But the unique Veo 3 can understand the raw pixel units of their videos and the sounds created with the clips automatically, for all Google.
Here is a sample of the form:
Veo 3 is likely possible Previous Deepmind work In “Video to Anto” Amnesty International. Last June, Deepmind revealed that he was developing artificial intelligence technology to create soundtracks for videos by training a model on a set of sounds and texts of dialogue as well as videos.
DeepMind will not say exactly where he got the content to train VEO 3, but YouTube is a strong possibility. Google YouTube and DeepMind owns previously Tell Techcrunch that Google models such as Veo “May” are trained in some YouTube materials.
To reduce the risk of deeply developing, DeepMind says it uses its watermark technology, Synthid, to include invisible signs in VEO 3 generation frames.
While companies like Google Pitch Veo 3 as strong creative tools, many artists are concerned about it – they threaten to raise the entire industries. 2024 Ticket Encouraged by the Animation Syndicate, the union that represents the animation in Hollywood and cartoonists, it is estimated that more than 100,000 films, television and animation in the United States will be disabled by artificial intelligence by 2026.
Google has also launched new possibilities for VEO 2, including a feature that allows users to provide typical images of characters, viewers, objects and patterns to improve consistency. Another VEO 2 can understand camera movements such as rotation, dolls and enlargement, and it allows users to add or erase beings from videos or expand clips frames, for example, converting them from image to scene.
Google says that all these new VEO 2 capabilities will come to its API platform in the coming weeks.