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

SAM2-ARAFNet: adapting SAM2 with an attention-enhanced residual ASPP fusion network for high-resolution remote sensing semantic segmentation.

Scientific reports·2026
Same author

Hierarchical Self-Distillation with Attention for Class-Imbalanced Acoustic Event Classification in Elevators.

Sensors (Basel, Switzerland)·2026
Same author

CAS-ViT: Convolutional Additive Self-Attention Vision Transformers for Efficient Mobile Applications.

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

Intracranial aneurysm segmentation with nnU-net: utilizing loss functions and automated vessel extraction.

Vessel plus·2026
Same author

SAMURAI: Motion-Aware Memory for Training-Free Visual Object Tracking With SAM 2.

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

GeoNR-PSW: Prompt-Aligned Localization Leveraging Ray-Traced 5G Channels and LLM Reasoning.

Sensors (Basel, Switzerland)·2025

Related Experiment Video

Updated: Oct 23, 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

710

IRFR-Net: Interactive Recursive Feature-Reshaping Network for Detecting Salient Objects in RGB-D Images.

Wujie Zhou, Qinling Guo, Jingsheng Lei

    IEEE Transactions on Neural Networks and Learning Systems
    |August 20, 2021
    PubMed
    Summary

    This study introduces the IRFR-Net, a novel dual-stream network for salient object detection (SOD) that effectively fuses RGB and depth data. The network achieves superior performance by leveraging gated attention and linear fusion for enhanced feature extraction and integration.

    More Related Videos

    Investigating Object Representations in the Macaque Dorsal Visual Stream Using Single-unit Recordings
    07:08

    Investigating Object Representations in the Macaque Dorsal Visual Stream Using Single-unit Recordings

    Published on: August 1, 2018

    8.4K
    Author Spotlight: Assessment of Visual Acuity in Central Vision Loss Through Motion-Based Peripheral Vision Testing
    06:25

    Author Spotlight: Assessment of Visual Acuity in Central Vision Loss Through Motion-Based Peripheral Vision Testing

    Published on: February 23, 2024

    778

    Related Experiment Videos

    Last Updated: Oct 23, 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

    710
    Investigating Object Representations in the Macaque Dorsal Visual Stream Using Single-unit Recordings
    07:08

    Investigating Object Representations in the Macaque Dorsal Visual Stream Using Single-unit Recordings

    Published on: August 1, 2018

    8.4K
    Author Spotlight: Assessment of Visual Acuity in Central Vision Loss Through Motion-Based Peripheral Vision Testing
    06:25

    Author Spotlight: Assessment of Visual Acuity in Central Vision Loss Through Motion-Based Peripheral Vision Testing

    Published on: February 23, 2024

    778

    Area of Science:

    • Computer Vision
    • Artificial Intelligence
    • Machine Learning

    Background:

    • Salient Object Detection (SOD) commonly utilizes Red-Green-Blue (RGB) images and depth maps.
    • Existing models often struggle to fully exploit complementary depth information for accurate SOD.
    • Attention mechanisms and linear fusion are established techniques for feature extraction and fusion in deep learning.

    Purpose of the Study:

    • To develop a novel dual-stream network, the Interactive Recursive Feature-Reshaping Network (IRFR-Net), for improved salient object detection using RGB-D data.
    • To effectively integrate complementary information from both RGB images and depth maps.
    • To enhance the highlighting of salient objects through advanced feature fusion techniques.

    Main Methods:

    • Constructed a dual-stream interactive recursive feature-reshaping network (IRFR-Net) combining gated attention and linear fusion.
    • Employed a context extraction module (CEM) for low-level depth foreground information and a gated attention fusion module (GAFM) for RGB-D data fusion.
    • Integrated adjacent depth information globally and utilized weighted atrous spatial pyramid pooling (WASPP) for multiscale depth feature extraction.
    • Fused global and local features using a bottom-up scheme.

    Main Results:

    • The proposed IRFR-Net effectively extracts and fuses complementary features from RGB and depth data.
    • The network demonstrates superior performance in highlighting salient objects compared to existing methods.
    • Comprehensive experiments on eight datasets show IRFR-Net outperforming 11 state-of-the-art (SOTA) RGB-D approaches.

    Conclusions:

    • The IRFR-Net architecture successfully leverages gated attention and linear fusion for robust salient object detection.
    • The proposed method significantly enhances the utilization of depth information in conjunction with RGB data.
    • IRFR-Net represents a significant advancement in RGB-D based salient object detection, achieving SOTA results.