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: May 2, 2026

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

1.3K

A multi-modal deep learning framework with GAN-based fusion for enhanced landslide detection.

R Srivats1, Deepika Roselind Johnson1, G Logeswari1

  • 1School of Computer Science and Engineering, Vellore Institute of Technology, Chennai Campus, Chennai, India.

Plos One
|April 30, 2026
PubMed
Summary

This study introduces a hybrid deep learning model for precise landslide detection and segmentation, outperforming existing methods. The framework fuses multiple Convolutional Neural Networks (CNNs) with Generative Adversarial Networks (GANs) for enhanced accuracy and rapid post-disaster mapping.

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 self-orchestrating physics-discovering neural architecture for adaptive and fault-resilient robotic motion control via RPA and digital twins.

Scientific reports·2026
Same author

Anal sphincter function in conservatively managed rectal intussusception at long-term follow-up: A prospective anal acoustic reflectometry study with comparison to healthy volunteer data.

Colorectal disease : the official journal of the Association of Coloproctology of Great Britain and Ireland·2026
Same author

YOLOv11-Lite architecture for wildlife detection from drone images.

Frontiers in artificial intelligence·2026
Same author

A federated deep learning approach for SDN security with quantum optimized feature selection and hybrid MSDC net architecture.

Scientific reports·2026
Same author

Hybrid Bayesian deep learning model for predicting urban heat island intensity in African cities.

Scientific reports·2025
Same author

CHASHNIt for enhancing skin disease classification using GAN augmented hybrid model with LIME and SHAP based XAI heatmaps.

Scientific reports·2025

Area of Science:

  • Geosciences and Remote Sensing
  • Artificial Intelligence
  • Computer Vision

Background:

  • Accurate landslide detection and segmentation are crucial for disaster management and risk assessment.
  • Traditional methods often struggle with complex terrain and varying image resolutions.
  • Existing deep learning models may lack the ability to capture both global context and fine spatial details.

Purpose of the Study:

  • To develop a hybrid deep learning framework for accurate landslide detection and segmentation.
  • To improve upon existing single-CNN and ensemble models through multi-backbone feature fusion and adversarial refinement.
  • To create a system capable of generating GIS-ready probability maps for rapid post-disaster decision support.

Main Methods:

  • Integration of four pre-trained Convolutional Neural Networks (CNNs): VGG16, DenseNet201, ResNet50, and InceptionV3.

Related Experiment Videos

Last Updated: May 2, 2026

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

1.3K
  • Implementation of a Generative Adversarial Network (GAN)-based adversarial refinement module for enhanced segmentation accuracy.
  • Multi-backbone feature fusion to capture global terrain context and fine-grained spatial details.
  • Evaluation on three benchmark datasets: CAS Landslide, MS2LandsNet, and GDCLD.

    • Main Results:

    • Achieved high F1-scores: 97.24% (CAS Landslide), 93.70% (MS2LandsNet), and 94.75% (GDCLD).
    • Demonstrated significant improvements over fusion baselines (1.4-2.9%) and single-CNN models (4-7%).
    • Showcased consistent Intersection over Union (IoU) gains and improved boundary delineation.
    • Exhibited strong generalization capabilities across varying resolutions, terrain types, and triggering mechanisms.

    Conclusions:

    • The proposed hybrid deep learning framework offers high accuracy and scalability for landslide segmentation.
    • The combination of multi-backbone feature fusion and adversarial refinement is effective for capturing complex landslide features.
    • The system's low-latency inference and GIS-ready outputs make it suitable for operational monitoring and rapid decision support in disaster scenarios.