Google may have just started rolling out its Veo generative AI to business customers, but the company is wasting no time getting the new version of the video tool to early testers. On Monday, Google announced a preview of Veo 2. According to the company, Veo 2 “understands the language of cinematography.” Basically, that means you can refer to a specific type of film, cinematic effect or lens when you tell the model.
Additionally, Google says the new model has a better understanding of real-world physics and human movement. Properly modeling people in motion is something that all production models strive to do. So the company’s claim that the Veo 2 is better when it comes to both those issues is noteworthy. Of course, the samples provided by the company are not enough to know for sure; the real test of the Veo 2’s capabilities will come when someone tells it to produce a video of a fitness trainer’s routine. Oh, and speaking of things video models struggle with, Google says the Veo will produce artifacts like extra fingers “gradually.”
Separately, Google is releasing an upgrade to Imagen 3. In its text-to-image model, the company says the latest version produces brighter and better-composed images. Additionally, it can render various art styles with great accuracy. At the same time, it is also better to follow the instructions more faithfully. Sticking quickly was an issue I emphasized when the company made Imagen 3 available to Google Cloud customers earlier this month, so if nothing else, Google knows the areas where its AI models need to work.
The Veo 2 will gradually roll out to Google Labs users in the US. Currently, Google will limit testers to producing up to eight seconds of video at 720p. In context, Sora can produce up to 20 seconds of 1080p video, though doing so requires a $200 per month ChatGPT Pro subscription. As for the latest improvements in Imagen 3, those are available to Google Labs users in more than 100 countries through ImageFX.
Source link
