A Lightweight Person Detector for Surveillance Footage Based on YOLOv8n
To enable person detection tasks in surveillance footage to be deployed on edge devices and their efficient performance in resource-constrained environments in real-time, a lightweight person detection model based on YOLOv8n was proposed. This model balances high accuracy with low computational cost...
Saved in:
| Main Authors: | , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
MDPI AG
2025-01-01
|
| Series: | Sensors |
| Subjects: | |
| Online Access: | https://www.mdpi.com/1424-8220/25/2/436 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1832587541503541248 |
|---|---|
| author | Qicheng Wang Guoqiang Feng Zongzhe Li |
| author_facet | Qicheng Wang Guoqiang Feng Zongzhe Li |
| author_sort | Qicheng Wang |
| collection | DOAJ |
| description | To enable person detection tasks in surveillance footage to be deployed on edge devices and their efficient performance in resource-constrained environments in real-time, a lightweight person detection model based on YOLOv8n was proposed. This model balances high accuracy with low computational cost and parameter size. First, the MSBlock module was introduced into YOLOv8n. Then, a series of modifications were made to the MSBlock structure. Next, a heterogeneous PAFPN with improved MSBlock was formed using heterogeneous convolution kernels. Finally, AKConv, a variable kernel convolution, was applied to further reduce the number of parameters and the computational cost while improving accuracy. A series of experiments demonstrated that, due to these improvements, the proposed lightweight model achieved a reduction of nearly 10% in parameter size and 5% in the floating-point computational cost compared to the original YOLOv8n. Additionally, on a custom surveillance dataset, the model shows a 1.4% improvement in mAP@0.5:0.95, and on a more complex subset of the PASVOC public dataset, the model achieved a 2.8% improvement in mAP@0.5 and a 1.2% improvement in mAP@0.5:0.95, proving the high accuracy and generalization ability of the improved lightweight model. |
| format | Article |
| id | doaj-art-fcb54adb19d943579647b83208a26320 |
| institution | Kabale University |
| issn | 1424-8220 |
| language | English |
| publishDate | 2025-01-01 |
| publisher | MDPI AG |
| record_format | Article |
| series | Sensors |
| spelling | doaj-art-fcb54adb19d943579647b83208a263202025-01-24T13:48:55ZengMDPI AGSensors1424-82202025-01-0125243610.3390/s25020436A Lightweight Person Detector for Surveillance Footage Based on YOLOv8nQicheng Wang0Guoqiang Feng1Zongzhe Li2Equipment Management and UAV Engineering School, Air Force Engineering University, Xi’an 710051, ChinaEquipment Management and UAV Engineering School, Air Force Engineering University, Xi’an 710051, ChinaEquipment Management and UAV Engineering School, Air Force Engineering University, Xi’an 710051, ChinaTo enable person detection tasks in surveillance footage to be deployed on edge devices and their efficient performance in resource-constrained environments in real-time, a lightweight person detection model based on YOLOv8n was proposed. This model balances high accuracy with low computational cost and parameter size. First, the MSBlock module was introduced into YOLOv8n. Then, a series of modifications were made to the MSBlock structure. Next, a heterogeneous PAFPN with improved MSBlock was formed using heterogeneous convolution kernels. Finally, AKConv, a variable kernel convolution, was applied to further reduce the number of parameters and the computational cost while improving accuracy. A series of experiments demonstrated that, due to these improvements, the proposed lightweight model achieved a reduction of nearly 10% in parameter size and 5% in the floating-point computational cost compared to the original YOLOv8n. Additionally, on a custom surveillance dataset, the model shows a 1.4% improvement in mAP@0.5:0.95, and on a more complex subset of the PASVOC public dataset, the model achieved a 2.8% improvement in mAP@0.5 and a 1.2% improvement in mAP@0.5:0.95, proving the high accuracy and generalization ability of the improved lightweight model.https://www.mdpi.com/1424-8220/25/2/436object detectionYOLOlightweight networkdeep learningsurveillance footage |
| spellingShingle | Qicheng Wang Guoqiang Feng Zongzhe Li A Lightweight Person Detector for Surveillance Footage Based on YOLOv8n Sensors object detection YOLO lightweight network deep learning surveillance footage |
| title | A Lightweight Person Detector for Surveillance Footage Based on YOLOv8n |
| title_full | A Lightweight Person Detector for Surveillance Footage Based on YOLOv8n |
| title_fullStr | A Lightweight Person Detector for Surveillance Footage Based on YOLOv8n |
| title_full_unstemmed | A Lightweight Person Detector for Surveillance Footage Based on YOLOv8n |
| title_short | A Lightweight Person Detector for Surveillance Footage Based on YOLOv8n |
| title_sort | lightweight person detector for surveillance footage based on yolov8n |
| topic | object detection YOLO lightweight network deep learning surveillance footage |
| url | https://www.mdpi.com/1424-8220/25/2/436 |
| work_keys_str_mv | AT qichengwang alightweightpersondetectorforsurveillancefootagebasedonyolov8n AT guoqiangfeng alightweightpersondetectorforsurveillancefootagebasedonyolov8n AT zongzheli alightweightpersondetectorforsurveillancefootagebasedonyolov8n AT qichengwang lightweightpersondetectorforsurveillancefootagebasedonyolov8n AT guoqiangfeng lightweightpersondetectorforsurveillancefootagebasedonyolov8n AT zongzheli lightweightpersondetectorforsurveillancefootagebasedonyolov8n |