Google has revealed the Veo generative artificial intelligence model, which can produce high-quality videos based on written requests from users, during the Google I/O 2024 Developers Conference.
Google claims that the Veo model has advanced capabilities in understanding natural language and visual cues, enabling the creation of any imaginable video by users.
Videos can be created using the Veo model lasting more than a minute, with a resolution of up to 1080 pixels. The model is also capable of understanding cinematic and visual techniques, such as the concept of a timeline, as indicated by Google.
Google collaborated with director Donald Glover and Gilga film studio to showcase the remarkable capabilities of the Veo model in simulating real-world physics, which can be seen in the promotional video released by the company on YouTube.
The Veo platform will be available today as part of Google’s VideoFX tool for some content creators, and will also be integrated into YouTube Shorts and other company’s products.
Google also announced the advanced Imagen 3 model for converting text into images, claiming it is the “highest quality” model in this field, offering an amazing level of detail and lifelike realistic images with fewer errors.
Google stated that Imagen 3 model now interacts better with texts and has become smarter in understanding details in long requests.
On the other hand, the tech community looks forward to exploring the new models offered by Google and comparing their performance with competing models from companies like OpenAI, such as the Sora model that converts texts into videos and the DALL-E 3 model that turns texts into images.
