File size: 1,069 Bytes
95e845d
 
97c412b
95e845d
97c412b
 
 
5e8a287
97c412b
32b95c5
97c412b
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
---
license: apache-2.0
pipeline_tag: text-to-image
---

# Show the Signal, Hide the Noise: Spectral Forcing for Pixel-Space Diffusion

This repository contains the weights for **Spectral Forcing**.

- **Paper:** arxiv.org/abs/2606.15236 
- **Github Repository:** https://github.com/WeichenFan/Spectral_Forcing

## Introduction
Spectral Forcing (SF) is a parameter-free, time-conditional low-pass operator that makes the coarse-to-fine structure of diffusion explicit at the input of a pixel-space model. It is based on the observation that under rectified-flow diffusion and natural-image power-law spectra, a moving frequency front separates a signal-bearing low-frequency region from a noise-dominated high-frequency region at each timestep, and that a standard denoiser wastes capacity outside this region by reproducing closed-form maps rather than modeling the data. SF imposes this boundary directly via a 2D-DCT mask whose cutoff expands with the diffusion time and becomes the identity at the data endpoint, showing the network the signal and hiding the noise.