Jove
Visualize
Contact Us
JoVE
x logofacebook logolinkedin logoyoutube logo
ABOUT JoVE
OverviewLeadershipBlogJoVE Help Center
AUTHORS
Publishing ProcessEditorial BoardScope & PoliciesPeer ReviewFAQSubmit
LIBRARIANS
TestimonialsSubscriptionsAccessResourcesLibrary Advisory BoardFAQ
RESEARCH
JoVE JournalMethods CollectionsJoVE Encyclopedia of ExperimentsArchive
EDUCATION
JoVE CoreJoVE BusinessJoVE Science EducationJoVE Lab ManualFaculty Resource CenterFaculty Site
Terms & Conditions of Use
Privacy Policy
Policies

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

A simulated experiment to explore robotic dialogue strategies for people with dementia.

Journal of rehabilitation and assistive technologies engineering·2022
Same author

Fast, low-memory detection and localization of large, polymorphic inversions from SNPs.

PeerJ·2022
Same author

Expression of IL-23/Th17 pathway in a murine model of Coxsackie virus B3-induced viral myocarditis.

Virology journal·2011
Same author

Therapeutic effect of carboxymethylated and quanternized chitosan on insulin resistance in high-fat-diet-induced rats and 3T3-L1 adipocytes.

Journal of biomaterials science. Polymer edition·2011
Same author

Role of reactive oxygen species in triptolide-induced apoptosis of renal tubular cells and renal injury in rats.

Journal of Huazhong University of Science and Technology. Medical sciences = Hua zhong ke ji da xue xue bao. Yi xue Ying De wen ban = Huazhong keji daxue xuebao. Yixue Yingdewen ban·2011
Same author

Chlorination and ortho-acetoxylation of 2-arylbenzoxazoles.

Organic & biomolecular chemistry·2011

Related Experiment Video

Updated: Nov 10, 2025

Author Spotlight: Enhancement of Salient Object Detection for Smart Grid Applications
03:31

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

Published on: December 15, 2023

753

MSDU-Net: A Multi-Scale Dilated U-Net for Blur Detection.

Xiao Xiao1, Fan Yang1, Amir Sadovnik2

  • 1School of Telecommunications Engineering, Xidian University, Xi'an 710071, China.

Sensors (Basel, Switzerland)
|April 3, 2021
PubMed
Summary

This study introduces MSDU-net, a novel multi-scale dilated convolutional neural network for improved image blur detection. The model effectively separates blurred and clear regions, outperforming existing methods.

Keywords:
U-shaped networkblur detectionimage segmentation

More Related Videos

Swin-PSAxialNet: An Efficient Multi-Organ Segmentation Technique
04:48

Swin-PSAxialNet: An Efficient Multi-Organ Segmentation Technique

Published on: July 5, 2024

630
Application of Deep Learning-Based Medical Image Segmentation via Orbital Computed Tomography
04:48

Application of Deep Learning-Based Medical Image Segmentation via Orbital Computed Tomography

Published on: November 30, 2022

3.1K

Related Experiment Videos

Last Updated: Nov 10, 2025

Author Spotlight: Enhancement of Salient Object Detection for Smart Grid Applications
03:31

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

Published on: December 15, 2023

753
Swin-PSAxialNet: An Efficient Multi-Organ Segmentation Technique
04:48

Swin-PSAxialNet: An Efficient Multi-Organ Segmentation Technique

Published on: July 5, 2024

630
Application of Deep Learning-Based Medical Image Segmentation via Orbital Computed Tomography
04:48

Application of Deep Learning-Based Medical Image Segmentation via Orbital Computed Tomography

Published on: November 30, 2022

3.1K

Area of Science:

  • Computer Vision
  • Deep Learning

Background:

  • Blur detection is crucial for computer vision tasks like object detection and face recognition.
  • Extracting and fusing multi-scale blur features remains a significant challenge.

Purpose of the Study:

  • To propose a novel deep learning model for effective blur detection.
  • To address the challenge of synchronous blur feature extraction and fusion.

Main Methods:

  • A multi-scale dilated convolutional neural network (MSDU-net) inspired by U-net architecture is proposed.
  • Multi-scale feature extractors with dilated convolutions are designed to capture texture information at various scales.
  • The U-shape architecture facilitates the fusion of multi-scale texture and semantic features.

Main Results:

  • Extensive experiments were conducted on two public benchmark datasets.
  • The proposed MSDU-net demonstrated superior performance compared to state-of-the-art blur detection approaches.

Conclusions:

  • MSDU-net effectively segments blurred and clear image regions.
  • The model offers a promising solution for enhancing image quality in computer vision applications.