Snapchat Research Introduces a New Text-to-Image Diffusion Model for Generative AI

Snapchat has just introduced a new text-to-image diffusion model for generative AI. The new approach presents a faster model for building visuals based on text queries. Most generative AI tools currently take around 30 seconds or so to generate such images. But Snap says that its new system is able to produce similar visuals within less than two seconds. It is undoubtedly an interesting development in the broader context of the generative AI process.

As explained by Snap:

“SnapFusion shortens the model runtime from a text input to image generation on mobile to under two seconds–the fastest time published to date by the academic community. Snap Research achieved this breakthrough by optimizing the network architecture and denoising process, making it incredibly efficient while maintaining image quality. So, now it’s possible to run the model to generate images based on text prompts, and get back crisp clear images in mere seconds on mobile rather than minutes or hours, as other research presents.”

Snapchat Research Introduces a New Text-to-Image Diffusion Model for Generative AI

Check Also: How to Remove Snapchat Filter from Saved Photo?

Snapchat also explained some examples of the visuals produced by the SnapFusion process. It looks much like the same type of generative AI pictures that you get from any other app. The result time is pretty faster, which Snap says could have a range of benefits.

Other than that Snap also notes that the new process could facilitate improved privacy. It also limits data sharing to third parties, while also reducing processing costs for developers.

It is worth mentioning here that the majority of its experiments were conducted on an iPhone Pro 14, which, in Snap’s own words ’has more computation power than many other phones’. It is uncertain whether other phones will also be quicker or not.

See Also: How to Fix Snapchat Can’t Load Images Problem?