Have a personal or library account? Click to login
Research on night vision pedestrian detection algorithm by incorporating attention mechanism Cover

Research on night vision pedestrian detection algorithm by incorporating attention mechanism

By: Jiaxue Yang and  Pingping Liu  
Open Access
|Dec 2025

Figures & Tables

Figure 1.

YOLO algorithm implementation flow
YOLO algorithm implementation flow

Figure 2.

Improved YOLOv8 network structure diagram
Improved YOLOv8 network structure diagram

Figure 3.

Partial data display of self-built night vision pedestrian data set
Partial data display of self-built night vision pedestrian data set

Figure 4.

LabelImg User Interface
LabelImg User Interface

Figure 5.

Content of the YOLO-formatted annotation txt file
Content of the YOLO-formatted annotation txt file

Figure 6.

TXT Format Structure of YOLOv8
TXT Format Structure of YOLOv8

Figure 7.

Training results prior to model optimization
Training results prior to model optimization

Figure 8.

Training results after model optimization
Training results after model optimization

Figure 9.

PR curve before algorithm improvement
PR curve before algorithm improvement

Figure 10.

PR curve after algorithm improvement
PR curve after algorithm improvement

Figure 11.

Image Detection Interface
Image Detection Interface

Figure 12.

Image Detection Interface
Image Detection Interface

Figure 13.

Video Detection Interface
Video Detection Interface

EXPERIMENTAL PLATFORM HARDWARE CONFIGURATION

NameRelated Configuration
Operating SystemWindows11
Processor11th Gen Intel(R) Core(TM) i7-11800H @ 2.30GHz 2.30 GHz
Memory16G
GPUNVIDIA GeForce RTX 3050
Deep LearningPytorch
Framework

COMPARISON OF ORIGINAL AND IMPROVED ALGORITHMS

AlgorithmPrecisionRecallmAP@0.5:0.95mAP@0.5F1 Score
YOLOv885.389.388.957.887.2
YOLOv8+CBAM87.384.488.9458.287.84

EXPERIMENTAL PARAMETER SETTINGS

ParameterParameter Setting
Number of training iterations50
Learning rate0.001
Batch size16
Image size640
OptimizerLion
Language: English
Page range: 22 - 33
Published on: Dec 31, 2025
In partnership with: Paradigm Publishing Services
Publication frequency: 4 issues per year

© 2025 Jiaxue Yang, Pingping Liu, published by Xi’an Technological University
This work is licensed under the Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 License.