Microsoft patents new way to train image generators using frequency noise
Microsoft describes a method for training diffusion models by adding noise to images in the frequency domain rather than the standard pixel domain. The process measures the natural variation in different frequency bands and applies specific amounts of noise to each, allowing the model to learn how to reverse that specific degradation.
In the filing’s own words
Certain aspects herein relate to iterative noising or denoising in diffusion models. According to one aspect herein, a forward process systematically introduces noise by transforming each training sample into a frequency representation and injecting frequency-specific perturbations based on a measured variance per band. According to another aspect herein, a reverse process starts from an initially noisy frequency-domain representation—also scaled to match known variances—and iteratively refines it toward a generated output. Both processes rely on frequency-aware transformations to modulate or restore data across multiple frequency bands, proceeding in opposite directions yet built on the same core principle of operating in the frequency domain using variance-based scaling.
What exactly they claim (claim 1)
1. A computer-implemented method of training a denoising model, the method comprising: receiving a training sample; converting the training sample into a frequency-domain representation; applying: a first frequency-specific degradation to a first frequency band of the frequency-domain representation based on a first variance of the first frequency band, and a second frequency-specific degradation to a second frequency band of the frequency-domain representation based on a second variance of the second frequency band, resulting in a degraded frequency-domain representation; and training the denoising model to reconstruct the frequency-domain representation based on the degraded frequency-domain representation, resulting in a trained denoising model.
Facts
- Company
- Microsoft (Microsoft Technology Licensing, LLC)
- Official title
- DIFFUSION MODEL TRAINING
- Publication number
- US20260300715A1
- Application number
- 19096437
- Filed
- Mar 31, 2025
- Published
- Oct 1, 2026
- Inventors
- 8
- Claims
- 20
Classified under
G06N 3/08 Physics & computingMore from Microsoft in AI & software
- Microsoft patents method to check AI answers against verified facts Oct 20, 2026
- Microsoft patents AI that presents PowerPoint slides for you Sep 3, 2026
- Microsoft patents method to check if LLM answers are based on real data Sep 10, 2026
- Microsoft patents system to blend multiple AI agents into one chat Oct 1, 2026
- Microsoft patents system that answers questions using your screen history Oct 15, 2026
Summaries are simplified from the filing’s own abstract and first claim. A patent filing is not a product announcement.
Patent data and drawings: U.S. Patent and Trademark Office (USPTO). Nebula is not affiliated with or endorsed by the USPTO.