Discover groundbreaking media creation models revealed by Google at the I/O Conference, transforming digital content creation forever.
At this year’s I/O developer conference, Google introduced its groundbreaking media creation models, with Veo 3 being the standout. Marking a significant advancement, Veo 3 is the first iteration capable of generating voice videos. For instance, it allows users to create videos featuring birdsong or urban scenes with ambient traffic noise. Google highlights that Veo 3 demonstrates excellence in real-world physics and lip-syncing. Currently, Veo 3 is accessible in the Gemini app for Gemini Ultra subscribers in the United States and through Vertex AI for enterprise users.
The introduction of Flow, Google’s latest AI-driven tool, enhances the video creation landscape by combining Veo, Imagen, and Gemini to produce cinematic clips and scenes. Users can articulate their desired outcomes in natural language, and Flow will generate content aligning with those descriptions. Initially, Flow is available exclusively to Google AI Pro and Ultra subscribers in the US, though Google intends to expand its availability globally soon.
Despite the launch of Veo 3, Google continues to support Veo 2. Users can utilize images of people, scenes, styles, and objects as reference points for content creation in Flow. Additional features include camera control capabilities, enabling users to rotate scenes and zoom in on specific objects. Furthermore, users have the flexibility to extend video frames from vertical to horizontal or add and remove elements within their content.
Google also unveiled its latest rendering model, Imagen 4, at the conference. Known for its ability to render intricate details such as complex fabrics and animal fur with outstanding clarity, Imagen 4 excels in creating both photorealistic and abstract images. Additionally, it offers superior typography rendering compared to its predecessors. Imagen 4 supports image creation in various aspect ratios and up to a 2K resolution. It is now available on the Gemini app, Vertex AI, and Workspace apps such as Docs and Slides. Notably, Google announced plans to release a version of Imagen 4 that is poised to be 10 times faster than Imagen 3 in the near future.
Yorum Yap