Consistency model mimics diffusion model denoising trajectory

← USPTO Patent Applications

CONSISTENCY MODEL WITH DENOISING ERROR

Application US20260093962A1 Kind: A1 Apr 02, 2026

Inventors

Noel Vouitsis, Rasa Hosseinzadeh, Brendan Leigh Ross, Victor Valentin Villecroze, Satya Krishna Gorti, Jesse Cole Cresswell, Gabriel Loaiza Ganem, Maksims Volkovs

Abstract

A consistency model is trained to mimic the output of a diffusion model at various points along the denoising trajectory. A trajectory of the diffusion model is determined by generating the data point with the diffusion model by sampling a noised data point and applying denoising steps of the diffusion model to obtain the denoised output. At each of the noise levels, the consistency model is applied to the corresponding data point to remove the remaining noise. The resulting data point from the consistency model is compared with the denoised output of the diffusion model. An error for the consistency model may then be determined based on the comparisons at the various points in the trajectory.

CPC Classifications

G06N 3/0475 G06F 17/13

Filing Date

2025-10-01

Application No.

19347043

View original document →