Search research articles

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

RGB-D Mirror Segmentation with Reliability-Guided Residual Correction.

Sensors (Basel, Switzerland)·2026

Same author

Aggregating Different Scales of Attention on Feature Variants for Tomato Leaf Disease Diagnosis from Image Data: A Transformer Driven Study.

Sensors (Basel, Switzerland)·2023

Same author

Multi-Stage Network for Event-Based Video Deblurring with Residual Hint Attention.

Sensors (Basel, Switzerland)·2023

Same author

Hint-Based Image Colorization Based on Hierarchical Vision Transformer.

Sensors (Basel, Switzerland)·2022

Same author

Global and Local Attention-Based Free-Form Image Inpainting.

Sensors (Basel, Switzerland)·2020

Same author

Deep Color Transfer for Color-Plus-Mono Dual Cameras.

Sensors (Basel, Switzerland)·2020

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: Aug 27, 2025

Author Spotlight: Enhancement of Salient Object Detection for Smart Grid Applications

Author Spotlight: Enhancement of Salient Object Detection for Smart Grid Applications

Published on: December 15, 2023

Multi-Scale Attention-Guided Non-Local Network for HDR Image Reconstruction.

Howoon Yoon¹, S M Nadim Uddin¹, Yong Ju Jung¹

¹School of Computing, Gachon University, Seongnam 13120, Korea.

Sensors (Basel, Switzerland)

|September 23, 2022

Summary

This study introduces MSANLnet, a novel network for high-dynamic-range (HDR) image reconstruction. It effectively reduces ghosting artifacts by using multi-scale attention and non-local fusion for improved HDR image quality.

Keywords:

deep learning high-dynamic-range imaging non-local means spatial attention

More Related Videos

Swin-PSAxialNet: An Efficient Multi-Organ Segmentation Technique

Swin-PSAxialNet: An Efficient Multi-Organ Segmentation Technique

Published on: July 5, 2024

Measuring Spatially- and Directionally-varying Light Scattering from Biological Material

Measuring Spatially- and Directionally-varying Light Scattering from Biological Material

Published on: May 20, 2013

Related Experiment Videos

Last Updated: Aug 27, 2025

Author Spotlight: Enhancement of Salient Object Detection for Smart Grid Applications

Author Spotlight: Enhancement of Salient Object Detection for Smart Grid Applications

Published on: December 15, 2023

Swin-PSAxialNet: An Efficient Multi-Organ Segmentation Technique

Swin-PSAxialNet: An Efficient Multi-Organ Segmentation Technique

Published on: July 5, 2024

Measuring Spatially- and Directionally-varying Light Scattering from Biological Material

Measuring Spatially- and Directionally-varying Light Scattering from Biological Material

Published on: May 20, 2013

Area of Science:

Computer Vision
Image Processing
Deep Learning

Background:

Current Convolutional Neural Network (CNN)-based methods for high-dynamic-range (HDR) image reconstruction often struggle with ghosting artifacts.
These artifacts arise from object or camera movement during the capture of multiple low-dynamic-range (LDR) images.

Purpose of the Study:

To propose an efficient HDR image reconstruction method that mitigates ghosting artifacts.
To enhance the quality of reconstructed HDR images, especially in the presence of motion.

Main Methods:

Introduced MSANLnet, a multi-scale attention-guided non-local network for HDR reconstruction.
Employed multi-scale spatial attention modules for implicit alignment of LDR image features.
Utilized non-local means-based fusion to reconstruct pixel intensity values by leveraging long-range dependencies.

Main Results:

MSANLnet demonstrated superior performance compared to state-of-the-art methods in quantitative evaluations.
Visual results confirmed the method's effectiveness in restoring saturated information and reducing ghosting artifacts caused by significant object movement.
Ablation studies validated the efficacy of the proposed architectural choices and modules.

Conclusions:

The proposed MSANLnet effectively addresses ghosting artifacts in HDR image reconstruction.
The multi-scale attention and non-local fusion approach leads to significant improvements in HDR image quality and robustness.