Google’s VEO 3 AI model creates videos

As part of this year’s ads at its I/O Deleper conference, Google unveiled the latest media generation models. The most prominent, perhaps, is Veo 3It is the first repetition of the model that can create videos with sounds. For example, a video clip of birds can be created with the sound of its singing, or the city street with traffic sounds in the background. Google says Veo 3 also excels in physics in the real world and in the synchronization of lips. Currently, the model is only available for Gemini Ultra subscribers in the United States within the Gemini app and Enterprise users on Vertex AI. It is also available in Flow, which is the new Google film manufacturing tool.

Wasting brings veo, Imagen GEMINI together to create clips and cinematic scenes. Users can describe the final output they want in natural, and the flow will go to work for them. The new tool will only be available to Google Ai Pro and Ultra in the United States at the present time, but Google says it will spread to more countries soon.

While the company released a new video generation model, it has not been abandoned by Veo 2 yet. Users will be able to give VEO 2 pictures to people, scenes, patterns and objects to use as a reference for their desired output to flow. They will have access to the camera controls that allow them to rotate the scenes and enlarge in specific flows of flow as well. In addition, they will be able to expand their frames from the image to the landscape if they want to add or remove objects from their videos.

Google also presented the latest photo generation models, Imagen 4, at this event. The company said that imagen 4 do accurate details such as complex fabrics and animal fur with “remarkable clarity” and outperform the generation of realistic and abstract images. It is also much better in providing printing from its ancestors and can create pictures of multi -phase ratios with decisions of up to 2K. Imagen 4 is now available via the Gemini and Vertex AI app and in the work space applications, including documents and slides. Google said it will also issue a version of IMAGEN 4 ten times faster than Imagen 3 “soon”.

Finally, to help people identify the content created by artificial intelligence, which has become more difficult these days, Google launched Synthid. It is a portal in which users can download part of the media that they believe can be created from artificial intelligence, and Google will determine whether it contains SurdishA watermark tool and identification of artificial intelligence art. Google has opened its watermark tool, but not all photo generators use it, so the gate is still able to determine all the images created by artificial intelligence.