Search research articles

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: May 31, 2026

End-To-End Deep Neural Network for Salient Object Detection in Complex Environments

End-To-End Deep Neural Network for Salient Object Detection in Complex Environments

Published on: December 15, 2023

A unified multi scale feature enhancement framework for remote sensing object detection.

Min Feng^1,2,3, Shuai Wang^4,5,6

¹School of Information Engineering, Shandong Youth University of Political Science, Jinan, 250103, China.

Scientific Reports

|May 29, 2026

Summary

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

The piRNA pathway is required for nucleopolyhedrovirus replication in Lepidoptera.

Insect science·2022

Same author

Efficacy of electro-acupuncture in postpartum with diastasis recti abdominis: A randomized controlled clinical trial.

Frontiers in public health·2022

Same author

Gas-liquid two-phase flow-based triboelectric nanogenerator with ultrahigh output power.

Science advances·2022

Same author

Silk sericin as building blocks of bioactive materials for advanced therapeutics.

Journal of controlled release : official journal of the Controlled Release Society·2022

Same author

Study on the Changes in Circadian Rhythm Before and After Treatment and the Influencing Factors in Patients with Depression.

Neuropsychiatric disease and treatment·2022

Same author

Apolipoprotein D facilitate the proliferation of BmNPV.

International journal of biological macromolecules·2022

Same journal

Peripheral B-cell receptor repertoire predicts immune-related adverse events following immune checkpoint inhibitor therapy in advanced renal cell carcinoma.

Scientific reports·2026

Same journal

Effects of black soldier fly (Hermetia illucens L.) larvae zoocompost on the mineral element content of blue honeysuckle berries.

Scientific reports·2026

Same journal

Investigation on absorption refrigeration performance of R1243zf with imidazolium ionic liquid as the working pairs.

Scientific reports·2026

Same journal

DeepTriage-CN: integrating clinical text with vital signs for emergency department admission prediction in an aging population.

Scientific reports·2026

Same journal

Gold nanoparticles as dual-action antiviral agents: disruption of SARS-CoV-2 viral envelopes and RNA integrity.

Scientific reports·2026

Same journal

Comparison of capillary microsampling and venous blood for multi-pathogen serosurveillance.

Scientific reports·2026

See all related articles

This summary is machine-generated.

This study introduces MSFE-YOLO, an efficient framework for small object detection in remote sensing. It enhances feature representation and fusion, improving accuracy for urban planning and environmental monitoring.

Area of Science:

Computer Vision
Remote Sensing Technology
Artificial Intelligence

Background:

Small object detection in remote sensing is crucial for applications like urban planning and disaster response.
Existing methods struggle with extreme scale variations and background clutter, often processing feature enhancement and multi-scale fusion separately.
Comprehensive contextual modeling in current approaches is computationally expensive.

Purpose of the Study:

To propose MSFE-YOLO, a unified and efficient multi-scale feature enhancement framework for small object detection in remote sensing.
To improve feature discrimination, bridge the semantic gap across scales, and aggregate multi-scale contextual information effectively.
To achieve a favorable accuracy-efficiency trade-off for practical remote sensing applications.

Main Methods:

Keywords:

Adaptive multi-scale correlation attention module Adaptive point upsampler Selective cross-attention fusion Small object detection in remote sensing You only look once version 8 small

Related Experiment Videos

Last Updated: May 31, 2026

End-To-End Deep Neural Network for Salient Object Detection in Complex Environments

End-To-End Deep Neural Network for Salient Object Detection in Complex Environments

Published on: December 15, 2023

Introduced an Adaptive Multi-scale Correlation Attention (AMCA) module in the backbone for improved feature discrimination and noise suppression.
Designed a Selective Cross-Attention Fusion (SCAF) module with an Adaptive Point Upsampler (APU) for content-aware feature reconstruction and fusion.
Integrated a Multi-scale Dilated Receptive Field (MDRF) module into the detection head for efficient multi-scale contextual information aggregation.

Main Results:

MSFE-YOLO demonstrated significant performance gains on the Remote Sensing Object Detection (RSOD) and Vehicle Detection in Aerial Imagery (VEDAI) datasets.
Achieved improvements of +4.6% mAP@50 and +8.9% mAP@50-95 on the RSOD dataset compared to the baseline.
While increasing parameters slightly (11.17M to 12.7M), the framework maintained acceptable inference speeds (reduced from 105.62 FPS to 74.56 FPS).

Conclusions:

The proposed MSFE-YOLO framework effectively enhances multi-scale representation and detection accuracy for small objects in remote sensing.
The integrated modules (AMCA, SCAF, APU, MDRF) efficiently address limitations of existing methods.
The results indicate a promising balance between improved accuracy and computational efficiency, suitable for real-world remote sensing tasks.