OpenAI launched text-to-video three months ago Sora, sparking widespread discussion among netizens, media and people in the industry.GoogleAt the 2024 I/O Developer Conference held today, a benchmark product was also launched- Veo,It can generate "high quality" videos with a length of more than 1 minute, a resolution of up to 1080P, and a variety of visual and cinematic styles.
According to Google's official press release, Veo has advanced understanding of natural language and can understand movie terms such as "time-lapse photography" and "aerial landscape".
Users can use text, image, or video prompts to guide their desired output, and Google says the resulting videos are "more coherent and consistent," with more realistic movements of people, animals, and objects throughout the shot.
Demis Hassabis, CEO of Google DeepMind, said at a media preview on Monday that video results can be refined with additional prompts, and Google is exploring more features that would enable Veo to produce storyboards and longer scenes.