Google has recently introduced new AI-based image upscaling technology that enhances the quality of low-resolution images.
In a post on Google’s AI blog titled “High Fidelity Image Generation Using Diffusion Model,” the researchers from Brain Team unveiled two diffusion models to generate high-fidelity images.
The research team presented two connected approaches that push the boundaries of the image synthesis quality for diffusion models. The two models are Super-Resolution via Repeated Refinements (SR3) and Cascaded Diffusion Models (CDM).
First is the SR3, which is a super-resolution diffusion model that takes as input a low-resolution image, and builds a corresponding high-resolution image from pure noise. Super-resolution has many applications that can range from restoring old family portraits to improving medical imaging systems.
According to the research team, this model is trained on an image corruption process in which noise is progressively added to a high-resolution image until only pure noise remains. It then learns to reverse this process, beginning from pure noise and progressively removing noise to reach a target distribution through the guidance of the input low-resolution image.
However, with large-scale training, SR3 achieves strong benchmark results on the super-resolution task for face and natural images when scaling to resolutions 4x–8x that of the input low-resolution image.
Meanwhile, after seeing the effectiveness of SR3, Google used these SR3 models for class-conditional image generation.
CDM is identified as a class-conditional diffusion model trained on ImageNet data to generate high-resolution natural images. “Since ImageNet is a difficult, high-entropy dataset, we built CDM as a cascade of multiple diffusion models,” Google said in the blog.
Furthermore, this cascade approach involves chaining together multiple generative models over several spatial resolutions: one diffusion model that generates data at a low resolution, followed by a sequence of SR3 super-resolution diffusion models that gradually increase the resolution of the generated image to the highest resolution.
“With SR3 and CDM, we have pushed the performance of diffusion models to state-of-the-art on super-resolution and class-conditional ImageNet generation benchmarks,” said Google.
Google will further test the limits of diffusion models for a wide variety of generative modeling problems.