18052658. DIFFUSION MODELS HAVING CONTINUOUS SCALING THROUGH PATCH-WISE IMAGE GENERATION simplified abstract (ADOBE INC.)
DIFFUSION MODELS HAVING CONTINUOUS SCALING THROUGH PATCH-WISE IMAGE GENERATION
Organization Name
Inventor(s)
Yinbo Chen of La Jolla CA (US)
Michaël Gharbi of San Francisco CA (US)
Oliver Wang of Seattle WA (US)
Richard Zhang of Burlingame CA (US)
Elya Shechtman of Seattle WA (US)
DIFFUSION MODELS HAVING CONTINUOUS SCALING THROUGH PATCH-WISE IMAGE GENERATION - A simplified explanation of the abstract
This abstract first appeared for US patent application 18052658 titled 'DIFFUSION MODELS HAVING CONTINUOUS SCALING THROUGH PATCH-WISE IMAGE GENERATION
Simplified Explanation
The patent application describes a method for generating an output image with semantic content by combining image patches based on a noise map and a global image code.
- Obtaining a noise map and a global image code from an original image
- Generating image patches using a diffusion model based on the noise map and global image code
- Combining the image patches to produce an output image with semantic content
Potential Applications
This technology could be applied in the fields of image processing, computer vision, and artificial intelligence for enhancing image quality and extracting semantic information from images.
Problems Solved
This technology solves the problem of efficiently generating output images with semantic content from original images by utilizing noise maps and global image codes.
Benefits
The benefits of this technology include improved image quality, enhanced semantic understanding of images, and the ability to generate output images with meaningful content.
Potential Commercial Applications
A potential commercial application of this technology could be in the development of image editing software, content creation tools, and image recognition systems.
Possible Prior Art
One possible prior art for this technology could be the use of image processing algorithms to enhance image quality and extract semantic information from images.
Unanswered Questions
How does this technology compare to existing image processing techniques?
This article does not provide a direct comparison to existing image processing techniques in terms of efficiency, accuracy, or performance.
What are the limitations of this technology in handling complex or large-scale images?
This article does not address the potential limitations of this technology when dealing with complex or large-scale images, such as computational resources required or processing time.
Original Abstract Submitted
Aspects of the methods, apparatus, non-transitory computer readable medium, and systems include obtaining a noise map and a global image code encoded from an original image and representing semantic content of the original image; generating a plurality of image patches based on the noise map and the global image code using a diffusion model; and combining the plurality of image patches to produce an output image including the semantic content.
