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

Broad-spectrum antifungal activity and genome-guided characterization of Paenibacillus polymyxa CACC1094 isolated from the bovine rumen.

PloS one·2026
Same author

Polygenic risk scores associate with asthma phenotypes and proteomic analyses implicate IL1R1 in two family-based studies.

medRxiv : the preprint server for health sciences·2026
Same author

Ultrasound-Based SECURED Protocol for Central Venous Catheterization: A Feasibility Study.

BioMed research international·2026
Same author

Comparing 2-Hz and 100-Hz Electroacupuncture for Postoperative Musculoskeletal Pain: Protocol for a 2 × 2 Prospective Randomized Crossover Trial.

JMIR research protocols·2026
Same author

Remimazolam-Induced Anaphylaxis After Spinal Anesthesia: A Case Report and Literature Review.

Journal of clinical medicine·2026
Same author

Preoperative and intraoperative oxygen consumption in patients undergoing lung resection surgery: a retrospective study using cardiopulmonary exercise test and intraoperative data.

BMC anesthesiology·2026

Related Experiment Video

Updated: Sep 26, 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

661

HISFCOS: Half-Inverted Stage Block for Efficient Object Detection Based on Deep Learning.

Beomyeon Hwang1, Sanghun Lee2, Seunghyun Lee2

  • 1Department of Plasma Bio Display, Kwangwoon University, Seoul 01897, Korea.

Journal of Imaging
|April 21, 2022
PubMed
Summary
This summary is machine-generated.

A new Half-Inverted Stage FCOS (HISFCOS) improves object detection accuracy by minimizing feature loss with a novel HIS block. This method enhances detection performance on benchmark datasets with comparable computational costs to existing models.

Keywords:
CNNFCOSHIS blockdeep learningfeature pyramidobject detection

More Related Videos

A Step-by-Step Implementation of DeepBehavior, Deep Learning Toolbox for Automated Behavior Analysis
05:41

A Step-by-Step Implementation of DeepBehavior, Deep Learning Toolbox for Automated Behavior Analysis

Published on: February 6, 2020

9.5K
A Swin Transformer-Based Model for Thyroid Nodule Detection in Ultrasound Images
04:23

A Swin Transformer-Based Model for Thyroid Nodule Detection in Ultrasound Images

Published on: April 21, 2023

2.0K

Related Experiment Videos

Last Updated: Sep 26, 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

661
A Step-by-Step Implementation of DeepBehavior, Deep Learning Toolbox for Automated Behavior Analysis
05:41

A Step-by-Step Implementation of DeepBehavior, Deep Learning Toolbox for Automated Behavior Analysis

Published on: February 6, 2020

9.5K
A Swin Transformer-Based Model for Thyroid Nodule Detection in Ultrasound Images
04:23

A Swin Transformer-Based Model for Thyroid Nodule Detection in Ultrasound Images

Published on: April 21, 2023

2.0K

Area of Science:

  • Computer Vision
  • Machine Learning
  • Deep Learning

Background:

  • Object detection is crucial for industrial applications.
  • Conventional Fully Convolutional One-Stage Detector (FCOS) methods suffer from low detection accuracy relative to their computational cost.
  • Low-level information loss in FCOS contributes to reduced accuracy.

Purpose of the Study:

  • To propose a Half-Inverted Stage FCOS (HISFCOS) that enhances detection accuracy while maintaining comparable computational costs to FCOS.
  • To address the issue of low-level information loss in conventional FCOS models.
  • To introduce a novel Half-Inverted Stage (HIS) block for improved feature extraction.

Main Methods:

  • Proposed a Half-Inverted Stage (HIS) block designed to extract spatial and channel information in parallel, minimizing feature loss.
  • Reconstructed the feature pyramid using the HIS block to improve low-level information representation.
  • Developed an improved detection head structure to reduce computational cost and complexity.
  • Trained and evaluated HISFCOS on PASCAL VOC and MSCOCO2017 datasets using Average Precision (AP) as the evaluation metric.

Main Results:

  • The HISFCOS model demonstrated improved detection accuracy, achieving a 3.0 AP increase on PASCAL VOC and a 1.5 AP increase on MSCOCO datasets compared to the conventional FCOS.
  • The number of parameters increased by only 0.5 M compared to the conventional method.
  • Ablation studies confirmed the effectiveness of the proposed HIS block and detection head structure.

Conclusions:

  • The proposed HISFCOS effectively improves object detection accuracy by mitigating low-level information loss through the HIS block and an optimized detection head.
  • HISFCOS offers a computationally efficient alternative to existing methods, achieving superior performance on benchmark datasets.
  • The study validates the contribution of the HIS block and refined detection head to enhanced object detection capabilities.