The latest research from Jasper signals a significant milestone for AI image generation capabilities and out-performs contributions from major players including MIT and Adobe

Jasper, the world’s leading AI marketing platform, released a research paper focused on Flash Diffusion, an innovative new distillation method developed by Jasper Research. Flash Diffusion is one of the fastest, most robust and versatile open-source distillation methods available for diffusion algorithms. Jasper’s results outperform the latest research on diffusion models from MIT, Adobe, and Bytedance.

Marketing Technology News: Tiger Pistol’s New Playbook Unlocks the Power of Vertical Video Advertising

The paper, “Flash Diffusion: Accelerating Any Conditional Diffusion Model for Few Steps Image Generation,” highlights traditional distillation methods that require extensive training time and computational resources. The new method identified by Jasper Research cuts down required training parameters by a minimum of 2.5x and up to 64x depending on the method, making training orders of magnitude faster. The new method also speeds up inference by 500%, leading to a 5x reduction in GPU inference costs and unlocking real-time applications such as text-to-image, in-painting and image upscaling. While previous methods are mostly closed-source and difficult to reproduce, Jasper open-sourced the codebase to showcase reproducibility. Overall, this new method enables faster training, lower overhead, and higher quality output.

“Marketing is inherently visual, and this field of research will make text-to-image generation faster, more versatile, and allow us to build new multimodal models that will support enterprise marketing initiatives for our customers,” said Suhail Nimji, Chief Strategy Officer. “Our research will continue to introduce state-of-the-art image manipulation capabilities for every visual asset needed in a marketer’s toolbox from website and product imagery to campaign and branding materials.”

Marketing Technology News: MarTech Interview with Baruch Toledano, Vice President and GM of Digital Marketing Solutions @ Similarweb

“The release of Flash Diffusion is advancing the boundaries of AI in image manipulation. I am vocal about my support for open source models and think our community will benefit from this application being hosted on Hugging Face for widespread use and adoption,” said Julien Chaumond, Hugging Face co-founder and CTO.

The Jasper-produced research will be hosted on Hugging Face as an open-source model publicly available to the research community and can be viewed HERE. To learn more about Flash Diffusion, read the research paper HERE. Led by Damien Henry, the Paris-based Jasper Research Lab is currently hiring for several technical and scientific positions, including backend engineers and a product manager.