Small target detection algorithm based on the fusion attention mechanism and multi-layer convolution.
In the realm of unmanned aerial vehicles, we proposed an enhanced small target detection algorithm, MGAC-YOLO, to address the challenges of missed detections and low accuracy associated with small target identification. Initially, we designed the MConv (Multi-layer Convolution) module to replace the...
Saved in:
| Main Authors: | , , , |
|---|---|
| Format: | Article |
| Language: | English |
| Published: |
Public Library of Science (PLoS)
2025-01-01
|
| Series: | PLoS ONE |
| Online Access: | https://doi.org/10.1371/journal.pone.0328003 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1849243382397796352 |
|---|---|
| author | Xiujing Li Haifei Zhang Yiliu Hang Hao Chen |
| author_facet | Xiujing Li Haifei Zhang Yiliu Hang Hao Chen |
| author_sort | Xiujing Li |
| collection | DOAJ |
| description | In the realm of unmanned aerial vehicles, we proposed an enhanced small target detection algorithm, MGAC-YOLO, to address the challenges of missed detections and low accuracy associated with small target identification. Initially, we designed the MConv (Multi-layer Convolution) module to replace the conventional Conv module within the backbone network, thereby augmenting the dimensionality of information capture and enhancing the detection performance for small targets. Subsequently, we harnessed the advantages of both attention mechanisms-GAM (Global Attention Mechanism) and CloAttention (Contextualized Local and Global Attention)-to create a GACAttention module that extracts small target features from both global and local perspectives, thereby enriching the network's focus on small target feature information and further enhancing its feature processing capabilities. Finally, we incorporated an additional small target detection layer to capture feature information at a shallower level, thereby reducing the likelihood of missed detections and bolstering the detection capabilities for small targets. Experimental results on the VisDrone2019 dataset demonstrate that the Precision, mAP50, and mAP50-95 of the MGAC-YOLO algorithm have improved by 5.3%, 6.3%, and 4.4%, respectively, in comparison to the baseline model YOLOv8s. Furthermore, when compared to other leading algorithms, the MGAC-YOLO algorithm has exhibited notable superiority. |
| format | Article |
| id | doaj-art-41901866aae344d4ab540dd6d82ca1ed |
| institution | Kabale University |
| issn | 1932-6203 |
| language | English |
| publishDate | 2025-01-01 |
| publisher | Public Library of Science (PLoS) |
| record_format | Article |
| series | PLoS ONE |
| spelling | doaj-art-41901866aae344d4ab540dd6d82ca1ed2025-08-20T03:59:30ZengPublic Library of Science (PLoS)PLoS ONE1932-62032025-01-01207e032800310.1371/journal.pone.0328003Small target detection algorithm based on the fusion attention mechanism and multi-layer convolution.Xiujing LiHaifei ZhangYiliu HangHao ChenIn the realm of unmanned aerial vehicles, we proposed an enhanced small target detection algorithm, MGAC-YOLO, to address the challenges of missed detections and low accuracy associated with small target identification. Initially, we designed the MConv (Multi-layer Convolution) module to replace the conventional Conv module within the backbone network, thereby augmenting the dimensionality of information capture and enhancing the detection performance for small targets. Subsequently, we harnessed the advantages of both attention mechanisms-GAM (Global Attention Mechanism) and CloAttention (Contextualized Local and Global Attention)-to create a GACAttention module that extracts small target features from both global and local perspectives, thereby enriching the network's focus on small target feature information and further enhancing its feature processing capabilities. Finally, we incorporated an additional small target detection layer to capture feature information at a shallower level, thereby reducing the likelihood of missed detections and bolstering the detection capabilities for small targets. Experimental results on the VisDrone2019 dataset demonstrate that the Precision, mAP50, and mAP50-95 of the MGAC-YOLO algorithm have improved by 5.3%, 6.3%, and 4.4%, respectively, in comparison to the baseline model YOLOv8s. Furthermore, when compared to other leading algorithms, the MGAC-YOLO algorithm has exhibited notable superiority.https://doi.org/10.1371/journal.pone.0328003 |
| spellingShingle | Xiujing Li Haifei Zhang Yiliu Hang Hao Chen Small target detection algorithm based on the fusion attention mechanism and multi-layer convolution. PLoS ONE |
| title | Small target detection algorithm based on the fusion attention mechanism and multi-layer convolution. |
| title_full | Small target detection algorithm based on the fusion attention mechanism and multi-layer convolution. |
| title_fullStr | Small target detection algorithm based on the fusion attention mechanism and multi-layer convolution. |
| title_full_unstemmed | Small target detection algorithm based on the fusion attention mechanism and multi-layer convolution. |
| title_short | Small target detection algorithm based on the fusion attention mechanism and multi-layer convolution. |
| title_sort | small target detection algorithm based on the fusion attention mechanism and multi layer convolution |
| url | https://doi.org/10.1371/journal.pone.0328003 |
| work_keys_str_mv | AT xiujingli smalltargetdetectionalgorithmbasedonthefusionattentionmechanismandmultilayerconvolution AT haifeizhang smalltargetdetectionalgorithmbasedonthefusionattentionmechanismandmultilayerconvolution AT yiliuhang smalltargetdetectionalgorithmbasedonthefusionattentionmechanismandmultilayerconvolution AT haochen smalltargetdetectionalgorithmbasedonthefusionattentionmechanismandmultilayerconvolution |