Want smarter insights in your inbox? Sign up for our weekly newsletters to get only what matters to enterprise AI, data, and security leaders. Subscribe Now
Google‘s family of image generation models is receiving some significant updates. The company announced Imagen 2 will receive text-to-live capabilities, transforming a text prompt into an animated image. Additionally, image editing within the AI model — inpainting, outpainting and digital watermarking — is now generally available.
Announced at the Google Cloud Next conference, Imagen 2’s text-to-live generates the equivalent of animated GIFs delivered initially at 24 frames per second, 360×640 pixel resolution, and a duration of four seconds. However, Google says there are planes for “continuous enhancements.”
“If you look at, instead of having a picture of an object, like a static picture of a car, you can see a short image like an animated moving vehicle. Many organizations, particularly in areas like media and advertising, are looking at it because it improves engagement with users,” Google Cloud CEO Thomas Kurian shared during a press briefing.
Google boasts Imagen 2 will be able to create images using a range of camera angles and motions “while supporting consistency over the entire sequence.” It also features safety filters and digital watermarks, addressing organizations’ two major concerns surrounding generative AI.
AI Scaling Hits Its Limits
Power caps, rising token costs, and inference delays are reshaping enterprise AI. Join our exclusive salon to discover how top teams are:
- Turning energy into a strategic advantage
- Architecting efficient inference for real throughput gains
- Unlocking competitive ROI with sustainable AI systems
Secure your spot to stay ahead: https://bit.ly/4mwGngO
As for the now-public image editing features, Imagen 2 can add or remove elements from a photo, similar to Adobe Photoshop’s generative fill or content-aware. Plus, expanding the borders of an image to provide a more expansive view is possible, too.
These updates are part of Google’s announcements surrounding Vertex AI, its fully managed cloud AI platform.
Imagen 2 launched in 2023. It’s a product of Google’s DeepMind, marketed as an AI that generates photorealistic, high-resolution and aesthetically pleasing images from natural language prompts. It’s in the same league as OpenAI’s DALL-E, Midjourney and Adobe Firefly and made specifically to help enterprises create images that match its brand guidelines and governance needs.
