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: Dec 8, 2025

Author Spotlight: Enhancement of Salient Object Detection for Smart Grid Applications

Author Spotlight: Enhancement of Salient Object Detection for Smart Grid Applications

Published on: December 15, 2023

Automatic image annotation method based on a convolutional neural network with threshold optimization.

Jianfang Cao^1,2, Aidi Zhao¹, Zibang Zhang¹

¹School of Computer Science & Technology, Taiyuan University of Science and Technology, Taiyuan, China.

|September 23, 2020

Summary

This summary is machine-generated.

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

Fully-gapped topological insulating states in a bilayer Bi(111) film grown on α-MoO<sub>3</sub>.

Nanoscale·2026

Same author

Dual-side snapshot light field imaging for accurate 3D volumetry of opaque and drifting underwater particles.

Optics letters·2026

Same author

Online visualization of small defects for a high-speed rotating object with single-pixel oblique detection.

Optics letters·2025

Same author

Decoding Molecular Conformations of Dimeric Non-Fullerene Acceptors for High-Sensitivity Near-Infrared Organic Photodetectors.

Angewandte Chemie (International ed. in English)·2025

Same author

Co and CoPc Molecular Kondo Box on Gold Surface.

Physical review letters·2025

Same author

Ultrafast and wide-spectral single-pixel tracking via composite spinning patterns: publisher's note.

Optics letters·2025

Same journal

Invaders taking over-Mollusc faunal change in volcanic barrier lakes of the Albertine Rift biodiversity hotspot.

PloS one·2026

Same journal

AI-driven molecular diversification and ligand-based optimization of macitentan derivatives targeting VEGFR1 and endothelin signaling pathways.

PloS one·2026

Same journal

Performance patterns and records in the world aquatics masters championships: Where do the most frequently represented nations among the top-ten masters swimmers come from?

PloS one·2026

Same journal

Modeling diurnal Temperature-Rainfall relationships under multicollinearity using PLS-SEM: A case study of Ghana.

PloS one·2026

Same journal

Organizational culture, social capital, and emergency capacity in primary healthcare institutions: A cross-sectional structural equation modeling study comparing ordinary and older communities.

PloS one·2026

Same journal

Impact of kidney function on the metabolome in the general population.

PloS one·2026

See all related articles

This study introduces CNN-THOP, a novel convolutional neural network with threshold optimization, to improve multilabel image annotation accuracy. The method effectively reduces overlabeling and downlabeling errors, enhancing annotation precision.

Area of Science:

Computer Vision
Machine Learning
Artificial Intelligence

Background:

Multilabel image annotation is crucial for image retrieval and understanding.
Existing methods often suffer from overlabeling and downlabeling errors.
Accurate annotation is essential for effective downstream applications.

Purpose of the Study:

To propose a novel Convolutional Neural Network with Threshold Optimization (CNN-THOP) model.
To address overlabeling and downlabeling issues in multilabel image annotation.
To enhance the accuracy and efficiency of image annotation processes.

Main Methods:

Developed CNN-THOP by integrating a threshold optimization algorithm with a Convolutional Neural Network (CNN).
Utilized batch normalization (BN) to accelerate model convergence and obtain prediction probabilities.

Related Experiment Videos

Last Updated: Dec 8, 2025

Author Spotlight: Enhancement of Salient Object Detection for Smart Grid Applications

Author Spotlight: Enhancement of Salient Object Detection for Smart Grid Applications

Published on: December 15, 2023

Implemented an optimal thresholding strategy for each label class based on prediction probabilities.

Main Results:

CNN-THOP demonstrated significant improvements in average precision across MIML, COREL5K, and MSRC datasets (27%-33% increase compared to MBRM).
Achieved a 3% increase in average recall rate on COREL5K and MSRC datasets compared to E2E-DCNN.
Reached a complete matching degree of 64.8% on the MIML dataset, indicating superior annotation precision.

Conclusions:

CNN-THOP effectively resolves overlabeling and downlabeling issues in multilabel image annotation.
The proposed model offers superior performance compared to existing methods like MBRM and E2E-DCNN.
CNN-THOP provides a robust and precise solution for automated image annotation tasks.