Google‘s got some big news. They’re upping their AI game. How, you ask? Well, they’re adding a new image generator to Google Bard. This was spilled in a yet-to-be-released changelog from January 18.
You’ve heard of tools like the Search Generative Experience (SGE) and Slides, right? Well, this is kinda like that. Users can whip up images by typing in a few descriptive words. And guess what’s powering this? Google’s Imagen family of models. These guys are famous for churning out top-notch, lifelike images.
Oh, and remember December? That’s when Google rolled out Imagen 2. This bad boy boasted a huge leap in image quality. It’s all about enterprise use cases, like making logos or images with text overlays.
Now, about this image generator in Google Bard. It’s gonna have safety features similar to SGE. Remember SGE? Introduced in October, it let users “draw” or “sketch” prompts right into the Search field. It could pop out four images at once.
And just like SGE, Google Bard ain’t gonna show results with famous folks or explicit stuff. SGE made sure to block harmful, misleading, or explicit content. Bard’s gonna do the same. It’ll also limit prompts featuring famous faces or photorealistic ones. This is according to 9to5Google. Plus, each image will have metadata labeling and invisible watermarking, thanks to Google’s SynthID.
You know Google Slides and Meet? They already let users generate images through Duet AI. Meet even lets you do it for custom backgrounds. Now, with Bard getting an image generator, AI-generated content is gonna be even more accessible across Google services.
And there’s more. Google’s adding more descriptions to image captions in Imagen 2’s training dataset. This means better image-caption understanding. It’ll lead to more accurate and contextually nuanced image generation.
But wait, there’s even more. Imagen 2 is tackling common issues faced by text-to-image tools. Think rendering realistic hands and human faces. It’s got a specialized image aesthetics model based on what humans prefer. This ensures the generated images align with what users like. All these advancements are boosting Imagen 2’s ability to generate high-quality, realistic images.
So, what’s the big picture? This move is part of Google’s wider strategy. They’re integrating advanced AI technologies across all their services. And that’s the scoop!