Google has started releasing private access to its AI production models Veo and Imagen 3. Starting today, customers of the company’s Vertex AI Google Cloud package can start using Veo to produce videos with messages and images. Then, starting next week, Google will make Imagen 3, its latest text-to-image framework, available to those same users.
With the release of Veo, Google says it is the first hyperscale cloud provider to offer a video-to-video model. Until then, OpenAI’s Sora model is still only available to select artists, academics and researchers — though that could soon change as the company teases 12 days of product demos starting December 5.
For Veo, Google says the model creates 1080p images that are “consistent and compact” and can work “in over a minute.” The tool is also able to work with both text and image guides. In the latter case, it is possible to use AI-generated or human-generated images as the starting point of the video.
If you look at the sample video shared by Google, it is clear that Veo, like all AI models, can struggle with cause and effect. For example, in a batch of roasted marshmallows, the treats don’t turn yellow and burn as they are exposed to the heat of the campfire flame. Artifacting is also a problem, as can be seen if you look closely at the hands in the concert photos.
As for Imagen 3, Google says the model produces “more realistic and high-quality images from simple text commands, surpassing previous versions of Imagen in detail, brightness, and artifact reduction.” Here again, however, you don’t have to look too hard to see that Google has a lot of work to do.
In the first example of a group of friends sitting on the trunk of a car, the actual information includes talking about “bright pictures,” but the subjects are clearly backlit. One might argue that the flash was used to create a strong background light, but if the idea behind the prompt was to create something representative of flash photography from the 1960s, this photo is not it.
Still, Google is determined to get more of its business customers using AI for productivity. Speaking about its research, the tech giant says that among companies that use productive AI in production, 86 percent report an increase in revenue. However, a recent Appen survey found the return on investment in AI projects decreased by 4.6 percent from 2023 to 2024.
If you purchase something through a link in this article, we may earn a commission.
Source link
