SCA-YOLO: 基于可变形卷积与上下文感知注意力的实时目标检测

邓立国; 沙文丹

doi:10.12052/gdutxb.240163

欢迎来到能源电力期刊网！

English Version 期刊登录

您当前的位置：

首页 >

文章列表页 >

SCA-YOLO: 基于可变形卷积与上下文感知注意力的实时目标检测

计算机科学与技术 | 更新时间：2026-05-27

- SCA-YOLO: 基于可变形卷积与上下文感知注意力的实时目标检测
- SCA-YOLO: Real-time Target Detection Based on Deformable Convolution and Context-aware Attention
- 广东工业大学学报 2026年43卷第2期页码：12-20
- 作者机构：
  
  1. 广东工业大学先进制造学院,广东,揭阳,515231
  2. 广东工业大学计算机学院,广东,广州,510006
- 作者简介：
  
  邓立国(1971–) ，男，副教授，博士，主要研究方向为数据挖掘、知识工程、大数据处理等，E-mail：denglg@gdut.edu.cn
- 基金信息：
  
  国家自然科学基金青年基金资助项目(62072120)
- DOI：10.12052/gdutxb.240163
  中图分类号： TP391
- 纸质出版：2026
- 稿件说明：
移动端阅览
邓立国, 沙文丹. SCA-YOLO: 基于可变形卷积与上下文感知注意力的实时目标检测[J]. 广东工业大学学报, 2026, 43(2): 12-20.
邓立国, 沙文丹. SCA-YOLO: 基于可变形卷积与上下文感知注意力的实时目标检测[J]. 广东工业大学学报, 2026, 43(2): 12-20. DOI： 10.12052/gdutxb.240163.

DOI： 10.12052/gdutxb.240163

摘要

YOLO(You Only Look Once)系列算法因其在目标检测中的实时性和准确性而广受赞誉。然而，其性能提升仍面临挑战：标准卷积受局部感受野限制，难以捕获全局上下文特征，影响复杂目标的检测精度；增大卷积核虽能增强特征提取，却显著增加计算成本，降低算法效率。针对上述问题，本文研究了SCA-YOLO模型，引入可变形通道融合模块(Alterable Channel-wise Fusion

C2fAK) 和上下文感知注意力机制(Context-aware Attention++

CAA++)模块提升性能。C2fAK模块结合可变形卷积与通道融合结构(Channel-wise Fusion

C2f)，提高特征表示能力并平衡计算开销；CAA++模块捕捉远距离上下文信息并减少通道冗余，进一步提升检测精度。实验表明，SCA-YOLO在多个数据集上的性能优于现有方法，展现了其在目标检测中的有效性和高效性。

Abstract

The YOLO (You Only Look Once) -based algorithms have been widely used for real-time object detection and achieved promising performance. However

this performance improvement still faces two challenges. First

standard convolutions with limited receptive fields are hard to capture global contextual features

reducing the detection accuracy of complex objects; Second

increasing the convolution kernel size can enhance feature extraction while the computational cost is significantly increased. To address these issues

this paper investigates the SCA-YOLO model by introducing the Alterable Channel-wise Fusion module (C2fAK) and the Context-Aware Attention++ (CAA++) module to enhance performance. The C2fAK module combines deformable convolution with the Channel-wise Fusion (C2f) structure to enhance feature representation capability while balancing computational overhead. The CAA++ module captures long-range contextual information and reduces channel redundancy

further improving detection accuracy. Experimental results show that the proposed SCA-YOLO outperforms existing methods on multiple datasets

demonstrating its effectiveness and efficiency in object detection.

关键词

Keywords

references

CHANG C M, LIOU Y D, HUANG Y C, et al . YOLO based deep learning on needle-type dashboard recognition for autopilot maneuvering system[J ] . Measurement and Control, 2022, 55(7-8): 567- 582.

ABBASI M, SHAHRAKI A, TAHERKORDI A. Deep learning for network traffic monitoring and analysis (NTMA): a survey[J]. Computer Communications, 2021, 170: 19-41.

XIE Y, ZHANG J, SHEN C, et al . Cotr: efficiently bridging CNN and transformer for 3D medical image segmentation[C ] //Medical Image Computing and Computer-Assisted Intervention–MICCAI 2021, 24th International Conference, Proceedings, Part III. Cham: Springer, 2021: 171-180.

LIU W, ANGUELOV D, ERHAN D, et al . SSD: single shot multibox detector[C ] //Computer Vision–ECCV 2016, 14th European Conference, Proceedings, Part I. Cham: Springer, 2016: 21-37.

REN S, HE K, GIRSHICK R, et al . Faster R-CNN: towards real-time object detection with region proposal networks[J ] . IEEE Transactions on Pattern Analysis and Machine Intelligence, 2016, 39(6): 1137-1149.

CARION N, MASSA F, SYNNAEVE G, et al . End-to-end object detection with transformers[C ] //European Conference on Computer Vision, 16th. Cham: Springer, 2020: 213-229.

ZHANG X, SONG Y, SONG T, et al . LDConv: linear deformable co nvolution for improving convolutional neural networks[J ] . Image and Vision Computing, 2024, 149: 105190.

REDMON J. You only look once: unified, real-time object detection[C]//Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition. New York: IEEE, 2016: 779-788.

WANG A, CHEN H, LIU L, et al . Yolov10: Real-time end-to-end object detection[J ] . Advances in Neural Information Processing Systems, 2024, 37: 107984-108011.

LI X S, TAN B H, YU R, et al . Small target detection algorithm for lightweight UAV aerial photography based on YOLOv5[J ] . Journal of Guangdong University of Technology, 2024, 41(3): 71-80.

DAI J, QI H, XIONG Y, et al . Deformable convolutional networks[C ] //Proceedings of the IEEE International Conference on Computer Vision. New York: IEEE, 2017: 764-773.

ZHAO S, ZHANG S, LU J, et al . A lightweight dead fish detection method based on deformable convolution and YOLOV4[J ] . Computers and Electronics in Agriculture, 2022, 198: 107098.

CHEN Y, DAI X, LIU M, et al . Dynamic convolution: attent ion over convolution kernels[C ] //Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. New York: IEEE, 2020: 11030-11039.

VASWANI A, SHAZEER N, PARMAR N, et al . Attention is all you need[C ] //Advances in Neural Information Processing Systems 30. Long Beach: NIPS, 2017: 5998-6008.

ZHANG H, DANA K, SHI J, et al . Context encoding for semantic segmentation[C ] //Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition. New York: IEEE, 2018: 7151-7160.

WOO S, PARK J, LEE J Y, et al . Cbam: convolutional block attention module[C ] //European Conference on Computer Vision, 15th. Cham: Springer, 2018: 3-19.

CAI X, LAI Q, WANG Y, et al . Poly kernel inception network for remote sensing detection[C ] //Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. New York: IEEE, 2024: 27706-27716.

LIN T Y, MAIRE M, BELONGIE S, et al . Microsoft coco: common objects in context[C ] //Computer Vision–ECCV 2014, 13th European Conference, Proceedings, Part V. Cham: Springer, 2014: 740-755.

TAN M, LE Q. Efficientnet: rethinking model scaling for convolutional neural networks[C]//International Conference on Machine Learning, 36th. Long Beach: PMLR, 2019: 6105-6114.

WANG N, ZHI M. A review of single-stage general object detection algorithms in deep learning[J]. Computer Science and Exploration, 2025, 19(5): 1115-1140.

浏览量

下载量

CSCD

文章被引用时，请邮件提醒。

提交

工具集

关联资源

面向电力场景图像的动态自适应增量目标检测DA-IOD算法

SR-Det：面向工业场景下细长和旋转目标的鲁棒检测

基于YOLOv4-MP的绝缘子爆裂缺陷检测方法

点云目标检测残差投票网络

基于特征融合的深度学习目标检测算法研究