Search research articles

ABOUT JoVE

Overview Leadership Blog JoVE Help Center

AUTHORS

Publishing Process Editorial Board Scope & Policies Peer Review FAQ Submit

LIBRARIANS

Testimonials Subscriptions Access Resources Library Advisory Board FAQ

RESEARCH

JoVE Journal Methods Collections JoVE Encyclopedia of Experiments Archive

EDUCATION

JoVE Core JoVE Business JoVE Science Education JoVE Lab Manual Faculty Resource Center Faculty Site

Terms & Conditions of Use

Search research articles

Related Experiment Video

Updated: Jun 29, 2025

Visualizing Visual Adaptation

Visualizing Visual Adaptation

Published on: April 24, 2017

DriftRec: Adapting Diffusion Models to Blind JPEG Restoration.

Simon Welker, Henry N Chapman, Timo Gerkmann

IEEE Transactions on Image Processing : a Publication of the IEEE Signal Processing Society

|April 5, 2024

Summary

This summary is machine-generated.

Related Concept Videos

You might also read

Related Articles

Articles linked to this work by shared authors, journal, and citation graph.

Sort by

Same author

High-throughput in situ single particle X-ray imaging of dehydrating viral capsids.

Light, science & applications·2026

Same author

Single-Particle X-ray Scattering Reveals a High Local Supersaturation of Precursors as the Origin of CoO Assembly Formation.

The journal of physical chemistry letters·2026

Same author

Statistical crystallography reveals an allosteric network in SARS-CoV-2 M<sup>pro</sup>.

Communications biology·2026

Same author

3D atomic structure determination with ultrashort-pulse MeV electron diffraction.

IUCrJ·2026

Same author

PEO-sheathed liquid jets increase sample delivery stability for serial femtosecond X-ray crystallography.

Scientific reports·2026

Same author

The long and short of it: Distinct natural crystal packing strategies of Cry toxins from Bacillus thuringiensis.

Structure (London, England : 1993)·2026

Same journal

Mask-guided Asymmetric Contrastive and Semantic Alignment for Unsupervised Person Re-Identification.

IEEE transactions on image processing : a publication of the IEEE Signal Processing Society·2026

Same journal

Hyperbolic Cycle Alignment for Infrared-Visible Image Fusion.

IEEE transactions on image processing : a publication of the IEEE Signal Processing Society·2026

Same journal

Learning Gaze Synthesizer via 3D-eye Controlled Diffusion and Cross-domain Feature Alignment.

IEEE transactions on image processing : a publication of the IEEE Signal Processing Society·2026

Same journal

Underlying Semantic Diffusion for Effective and Efficient In-Context Learning.

IEEE transactions on image processing : a publication of the IEEE Signal Processing Society·2026

Same journal

DiffRES: Unleashing Text-to-Image Diffusion Models for Generative Referring Expression Segmentation without Information Leakage.

IEEE transactions on image processing : a publication of the IEEE Signal Processing Society·2026

Same journal

Location Matters: Frequency-Spatial Dual Space Adaptation for Cross-Domain Few-Shot Segmentation.

IEEE transactions on image processing : a publication of the IEEE Signal Processing Society·2026

See all related articles

DriftRec uses diffusion models to restore highly compressed JPEG images, producing sharper results than other methods. This blind restoration technique requires minimal training data and generalizes well to various compression scenarios.

Area of Science:

Computer Vision
Machine Learning
Image Processing

Background:

JPEG compression significantly degrades image quality, especially at high compression levels.
Existing blind JPEG restoration methods often produce blurry outputs and struggle with diverse compression artifacts.

Purpose of the Study:

To develop a novel diffusion model-based method for high-fidelity blind JPEG restoration.
To improve image sharpness and distributional accuracy compared to existing techniques.

Main Methods:

Proposed DriftRec, a diffusion model adapted via a modified forward stochastic differential equation for image restoration.
Utilized the proximity of clean and corrupted image distributions to a learned prior, rather than a standard Gaussian prior.
Trained on clean/corrupted image pairs without prior knowledge of the compression operation.

More Related Videos

Integrating Visual Psychophysical Assays within a Y-Maze to Isolate the Role that Visual Features Play in Navigational Decisions

Integrating Visual Psychophysical Assays within a Y-Maze to Isolate the Role that Visual Features Play in Navigational Decisions

Published on: May 2, 2019

Fluorescence Recovery after Merging a Droplet to Measure the Two-dimensional Diffusion of a Phospholipid Monolayer

Fluorescence Recovery after Merging a Droplet to Measure the Two-dimensional Diffusion of a Phospholipid Monolayer

Published on: October 15, 2015

Related Experiment Videos

Last Updated: Jun 29, 2025

Visualizing Visual Adaptation

Visualizing Visual Adaptation

Published on: April 24, 2017

Integrating Visual Psychophysical Assays within a Y-Maze to Isolate the Role that Visual Features Play in Navigational Decisions

Integrating Visual Psychophysical Assays within a Y-Maze to Isolate the Role that Visual Features Play in Navigational Decisions

Published on: May 2, 2019

Fluorescence Recovery after Merging a Droplet to Measure the Two-dimensional Diffusion of a Phospholipid Monolayer

Fluorescence Recovery after Merging a Droplet to Measure the Two-dimensional Diffusion of a Phospholipid Monolayer

Published on: October 15, 2015

Main Results:

DriftRec outperformed L2 regression baselines and state-of-the-art methods in JPEG restoration.
The method successfully avoided generating blurry images, preserving image distribution more faithfully.
Achieved effective restoration with low noise levels and fewer sampling steps.

Conclusions:

DriftRec offers a robust and versatile solution for blind JPEG restoration, particularly at high compression levels.
The approach demonstrates strong generalization capabilities to unaligned double compression and real-world online JPEGs.
This work highlights the potential of tailored diffusion models for image restoration tasks.