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 Experiment Video

Updated: Nov 30, 2025

Deep Neural Networks for Image-Based Dietary Assessment
13:19

Deep Neural Networks for Image-Based Dietary Assessment

Published on: March 13, 2021

9.7K

DLNR-SIQA: Deep Learning-Based No-Reference Stitched Image Quality Assessment.

Hayat Ullah1, Muhammad Irfan1, Kyungjin Han1

  • 1Mixed Reality and Interaction Lab, Department of Software, Sejong University, Seoul 143-747, Korea.

Sensors (Basel, Switzerland)
|November 17, 2020
PubMed
Summary

This study introduces a novel deep learning approach for assessing the quality of stitched images in immersive content. The method effectively localizes and segments stitching errors, improving quality evaluation for virtual and augmented reality applications.

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

Barium titanate-based perovskites for multifunctional applications: comprehensive computational and experimental insights into electronic, thermoelectric, phonon dispersion and optical characteristics.

RSC advances·2026
Same author

First-Principles Design of Room Temperature Ferromagnetic Metallic Rare-Earth Zintl Compounds AB<sub>2</sub>C<sub>2</sub> (A = Ce, Pr, Nd; B = Li; C = Sb) for Next-Generation Spintronic and Magneto-Electronic Applications.

ChemistryOpen·2026
Same author

A Robust Multivariate Thresholding Function for Sparse and Biomedical Signal Reconstruction.

Sensors (Basel, Switzerland)·2026
Same author

Burden of Secondary Hyperparathyroidism Among Patients on Hemodialysis: A Cross-Sectional Study.

Cureus·2026
Same author

New Insights Into Curcumin Behavior After Its Biosynthesis in Engineered Escherichia coli Through Spectroscopic and Spatial Analysis.

Phytochemical analysis : PCA·2025
Same author

A Deep Learning Pipeline for Epilepsy Genomic Analysis Using GPT-2 XL and NVIDIA H100.

ArXiv·2025

Area of Science:

  • Computer Vision
  • Artificial Intelligence
  • Immersive Media Technology

Background:

  • Advancements in virtual reality (VR) and augmented reality (AR) drive demand for high-quality immersive content.
  • Deep learning shows promise for image quality assessment, but existing methods struggle with panoramic stitching errors.

Purpose of the Study:

  • To develop a novel deep learning-based No-Reference Stitched Image Quality Assessment (DLNR-SIQA) method.
  • To address limitations of current methods in localizing and segmenting stitching errors in immersive content.

Main Methods:

  • A three-fold approach utilizing Mask R-CNN (Regional Convolutional Neural Network) for error localization and segmentation.
  • Fine-tuning Mask R-CNN on annotated stitching error datasets.
  • Quality assessment based on segmented distorted regions.
Keywords:
computer visiondeep learningimage quality assessmentimage segmentationimmersive contents

More Related Videos

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

3.2K

Related Experiment Videos

Last Updated: Nov 30, 2025

Deep Neural Networks for Image-Based Dietary Assessment
13:19

Deep Neural Networks for Image-Based Dietary Assessment

Published on: March 13, 2021

9.7K
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

3.2K

Main Results:

  • The proposed DLNR-SIQA method successfully segments and localizes stitching errors in immersive content.
  • Quantitative and qualitative comparisons show superior performance over existing state-of-the-art techniques.
  • The method outperforms both full reference (FR-IQA) and no reference (NR-IQA) methods.

Conclusions:

  • The novel DLNR-SIQA approach provides efficient and accurate quality assessment for immersive content.
  • This method enhances the quality evaluation of panoramic images by focusing on localized stitching defects.
  • The findings contribute to improving the production and consumption of high-fidelity VR/AR experiences.