Related Concept Videos
Difference from Background: Limit of Detection
The LOD indicates the presence or absence...
Infrared (IR) Spectroscopy: Overview
Different compounds display unique properties due to their...
Light Acquisition
IR Spectrometers
Detection of Black Holes
Their closest cousins are neutron stars, which are composed almost entirely of neutrons packed against each other, making them extremely dense. A neutron star has the same mass as the Sun but its diameter is only a few kilometers. Therefore, the escape velocity from their surface is close to the speed of light.
Not until the 1960s, when the first neutron...
Gas Chromatography: Types of Detectors-II
You might also read
Related Articles
Articles linked to this work by shared authors, journal, and citation graph.
Psychological Impact and Workload of COVID-19 on Healthcare Workers in China During the Early Time of the Pandemic: A Cross-sectional Study.
Manipulation of Band Alignment in Two-Dimensional Vertical WSe<sub>2</sub>/BA<sub>2</sub>PbI<sub>4</sub> Ruddlesden-Popper Perovskite Heterojunctions via Defect Engineering.
Efficient and Stable Perovskite Solar Cells via CsPF<sub>6</sub> Passivation of Perovskite Film Defects.
Related Experiment Video
Updated: Aug 2, 2025

Tracking Infiltration Front Depth Using Time-lapse Multi-offset Gathers Collected with Array Antenna Ground Penetrating Radar
Published on: May 1, 2018
Improved YOLOv5 infrared tank target detection method under ground background.
Chao Liang1,2, Zhengang Yan3, Meng Ren3
1School of Artificial Intelligence, Xidian University, Xi'an, 710071, China. 1025743995@qq.com.
This study introduces a novel YOLOv5s-THSE model to enhance infrared tank detection accuracy. The improved model effectively suppresses complex backgrounds and boosts detection performance for ground targets.
Area of Science:
- Computer Vision
- Artificial Intelligence
- Infrared Imaging Technology
Background:
- Infrared seeker detection precision is crucial for guidance systems.
- Challenges in detecting ground tank targets include scale variations, complex backgrounds, and subtle infrared characteristics.
- Existing methods struggle with low target detection accuracy in challenging infrared imaging scenarios.
Purpose of the Study:
- To develop an advanced You Only Look Once, Transform Head Squeeze-and-Excitation (YOLOv5s-THSE) model for improved infrared tank detection.
- To enhance the extraction of target features and suppress complex ground backgrounds.
- To increase the accuracy and stability of detecting small and inconspicuous infrared targets.
Main Methods:
- Integration of a multi-head attention mechanism into the backbone and neck of the YOLOv5s network.
- Incorporation of a Cross Stage Partial, Squeeze-and-Exclusion module in the network's neck.
- Introduction of a small object detection head and utilization of the CIoU loss function.
Main Results:
- The proposed YOLOv5s-THSE model demonstrates superior performance in detecting infrared tank targets against complex ground backgrounds.
- Effective suppression of background noise and enhanced focus on target features were achieved.
- Significant improvements in detection accuracy and training stability were observed compared to baseline YOLOv5s and other variants.
Conclusions:
- The YOLOv5s-THSE model offers a robust solution for enhancing infrared tank detection capabilities.
- The applied attention mechanisms and specialized modules effectively address challenges posed by complex environments.
- This research contributes to advancing the precision of infrared guidance systems through improved target recognition.

