基于改进 YOLOv5s 的小人脸检测

周丽芳; 胡振; 刘波

您当前的位置：

首页 >

文章列表页 >

基于改进 YOLOv5s 的小人脸检测

更新时间：2024-11-05

- 基于改进 YOLOv5s 的小人脸检测
- Small Face Detection Based on Improved YOLOv5s Network
- 智能科学与技术学报 2024年页码：1-10
- 作者机构：
  
  重庆邮电大学软件工程学院，重庆 400065
- 作者简介：
  
  周丽芳，zhoulf@cqupt.edu.cn
- 基金信息：
  
  国家自然科学基金项目(62276039;62221005);重庆市教委科技攻坚计划项目(KJZD-K201900601;KJZDK20220606)
- DOI：
  中图分类号： TP311.5
- 收稿：2024-03-27，
  
  修回：2024-09-27，
  
  网络首发：2024-10-04，
- 稿件说明：
移动端阅览
周丽芳,胡振,刘波.基于改进 YOLOv5s 的小人脸检测[J].智能科学与技术学报,

.Small Face Detection Based on Improved YOLOv5s Network[J].Chinese Journal of Intelligent Science and Technology,
周丽芳,胡振,刘波.基于改进 YOLOv5s 的小人脸检测[J].智能科学与技术学报, DOI：10.11959/j.issn.2096-6652.xxxxxx.

.Small Face Detection Based on Improved YOLOv5s Network[J].Chinese Journal of Intelligent Science and Technology, DOI：10.11959/j.issn.2096-6652.xxxxxx.

摘要

目前在复杂的真实应用场景中，小人脸检测面临着诸多挑战，如尺度、光照剧烈变化、计算量大和精度较低等问题。针对现有模型中小人脸容易被忽视的问题，本文提出了一种基于卷积核注意力机制的小人脸检测模型SK-YOLOv5s。首先，设计一种小人脸增强模块 SFEM，将浅层特征与深层特征进行融合与上采样，提高了小人脸特征图的分辨率，从而增强其特征；其次，在模型中引入 SKNet 注意力机制，能够多尺度自适应调节感受野大小，提高小人脸的检测率；最后采用 EIoU 作为目标损失函数，直接优化预测框与真实框的宽高差异，提高小人脸检测的精度和稳定性，相较于原有方法，所提方法在 WIDER Face 数据集中平均检测准确率提升了 3.1%。实验结果证明了该模型在真实场景下进行小人脸检测的可行性。

Abstract

In the complex real-world applications

the task of detecting small faces encounters numerous challenges

which include issues related to scale variation

abrupt lighting changes and suboptimal precision. In order to address the concerns of overlooking small facial structures within existing models

this study introduces a novel approach termed SK-YOLOv5s which is based on convolutional kernel attention mechanism. Firstly

we proposal a small face enhancement module to fuse multi-layer features

which can enhance the resolution of small faces and improve its features.Subsequently

we incorporates SKNet attention mechanism into the model to adaptively adjust receptive field sizes across multiple scales so that the detection efficacy of small face can be enhanced. Finally

EIoU is utilized as the loss function

which facilitate a direct optimization of the width and height discrepancies between predicted and actual bounding boxes. This approach enhances the precision and stability of small face detection. The performance on the WIDER Face dataset demonstrate a mean detection accuracy improvement of 3.1% over prevailing methods. The experimental results demonstrate the model's viability for detecting small faces in real-life scenarios.

关键词

Keywords

references

邱天衡,王玲,王鹏,等.基于改进YOLOv5的目标检测算法研究[J].计算机工程与应用, 2022, 58(13):11.DOI:10.3778/j.issn.1002-8331.2202-0093.

郭凡, 卢铉宇, 李嘉怡, 王红梅.基于深度神经网络的目标跟踪算法综述[J].航空兵器, 2024, 31(1).DOI:10.12132/ISSN.1673-5048.2022.0226.

PANTIC M, ROTHKRANTZ L J M. Automatic analysis of facial expressions: The state of the art[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2000, 22(12): 1424- 1445

KUMAR N, BERG A C, BELHUMEUR P N, et al. Attribute and simile classifiers for face verification[C]. IEEE International Conference on Computer Vision, Kyoto, Japan, 2009: 365- 372.

SCHAPIRE R E. The strength of weak learnability[J]. Machine Learning, 1990, 5(2): 197-227.

SCHAPIRE R, FREUND Y. A decision-theoretic generalization of on-line learning and an application to boosting[C]. Second European Conference on Computational Learning Theory, Barcelona , Spain, 1995: 23-37.

LIENHART R, MAYDT J. An extended set of haar-like features for rapid object detection[C]. IEEE Conference on Image Processing, Rochester, USA, 2002, 1: I-I.

MITA T, KANEKO T, HORI O. Joint haar-like features for face detection[C]. IEEE Conference on Computer Vision, Beijing, China, 2005, 2: 1619-1626.

VIOLA P, JONES M. Rapid object detection using a boosted cascade of simple features[C]. IEEE Conference on Computer Vision and Pattern Recognition, Kauai, USA, 2001, 1: I-I.

REDMON J, FARHADI A. YOLO9000: better, faster, stronger[C]. IEEE Conference on Computer Vision and Pattern Recognition, Honolulu, USA, 2017: 7263-7271.

Babiloni F , Marras I , Kokkinos F ,et al.Poly-NL: Linear Complexity Non-local Layers with Polynomials[J]. 2021.DOI:10.1109/ICCV48922.2021.01035.

Wang, Chien-Yao, I-Hau Yeh, and Hong-Yuan Mark Liao. "YOLOv9: Learning What You Want to Learn Using Programmable Gradient Information." arxiv preprint ar**v:2402.13616 (2024).

Wang, Chien Yao , A. Bochkovskiy , and H. Y. M. Liao . "YOLOv7: Trainable bag-of-freebies sets new state-of-the-art for real-time object detectors." arXiv arXiv, 2022

ZHANG S, ZHU X, LEI Z, et al. S3fd: single shot scale-invariant face detector[C]. IEEE Conference on Computer Vision, Venice, Italy, 2017: 192-201.

HU P, RAMANAN D. Finding tiny faces[C]. IEEE Conference on Computer Vision and Pattern Recognition, Honolulu, USA, 2017: 951-959.

Deng, Jiankang, et al. "Retinaface: Single-stage dense face localisation in the wild." arxiv preprint ar**v:1905.00641 (2019).

ZHANG B, LI J, WANG Y, et al. Asfd: automatic and scalable face detector[EB/OL]. (2020-3- 30) [2023-3-15]. https://arxiv.org/abs/2003.11228.

YANG S, LUO P, LOY C C, et al. Wider face: a face detection benchmark[C]. IEEE Conference on Computer Vision and Pattern Recognition, Las Vegas, USA, 2016: 5525–5533.

Zhu, Yanjia, et al. "Tinaface: Strong but simple baseline for face detection." ar**v preprint ar**v:2011.13183 (2020).

HE K, ZHANG X, REN S, et al. Deep residual learning for image recognition[C]. IEEE Conference on Computer Vision and Pattern Recognition, Las Vegas, USA, 2016: 770-778.

Wang J , Yuan Y , Yu G .Face Attention Network: An Effective Face Detector for the Occluded Faces[J]. 2017.DOI:10.48550/arXiv.1711.07246.

SUN X, WU P, HOI S C H. Face detection using deep learning: an improved faster rcnn approach[J]. Neurocomputing, 2018, 299: 42-50.

田庆, 胡蓉, 李佐勇, 等. 基于SE-YOLOv5s的绝缘子检测[J]. 智能科学与技术学报, 2021, 3(3): 312-321.

谢椿辉,吴金明,徐怀宇.改进YOLOv5的无人机影像小目标检测算法[J].计算机工程与应用, 2023, 59(9):9.DOI:10.3778/j.issn.1002-8331.2212-0336.

ZHENG Z, WANG P, LIU W, et al. Distance-iou loss: faster and better learning for bounding box regression[C]. AAAI Conference on Artificial Intelligence, New York, USA, 2020, 34(07): 12993-13000.

ZHANG Y F, REN W, ZHANG Z, et al. Focal and efficient iou loss for accurate bounding box regression[J]. Neurocomputing, 2022, 506: 146-157.

Zhang K , Zhang Z , Li Z ,et al.Joint Face Detection and Alignment Using Multitask Cascaded Convolutional Networks[J].IEEE Signal Processing Letters, 2016, 23(10):1499-1503.DOI:10.1109/LSP.2016.2603342.

LI X, WANG W, HU X, et al. Selective kernel networks[C]. IEEE Conference on Computer Vision and Pattern Recognition, Long Beach, USA, 2019: 510-519.

Albiero V, Chen X, Yin X, et al. img2pose: Face alignment and detection via 6dof, face pose estimation[C]//Proceedings of the IEEE/CVF conference on computer vision and pattern recognition. 2021: 7617-7627

Qi D , Tan W , Yao Q ,et al.YOLO5Face: Why Reinventing a Face Detector[J]. 2021.DOI:10.48550/arXiv.2105.12931.

Guo J , Deng J , Lattas A ,et al.Sample and Computation Redistribution for Efficient Face Detection[J]. 2021.DOI:10.48550/arXiv.2105.04714.

Liu Y, Wang F, Deng J, et al. Mogface:Towards a deeper appreciation on face detection[C]//Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition. 2022: 4093-4102.

Xu Y , Yan W , Yang G ,et al.CenterFace: Joint Face Detection and Alignment Using Face as Point[J].Scientific Programming, 2020, 2020:1-8.DOI:10.1155/2020/7845384.

Yoo Y J , Han D , Yun S .EXTD: Extremely Tiny Face Detector via Iterative Filter Reuse[J]. 2019.DOI:10.48550/arXiv.1906.06579

高强,唐福兴,李栋,et al.基于改进YOLOv5的密集场景行人检测方法研究[J].国外电子测量技术, 2023, 42(4):125-130.

浏览量

166

下载量

CSCD

文章被引用时，请邮件提醒。

提交

工具集

关联资源

基于改进YOLOv5s的小人脸检测

小样本细粒度图像分类的Mamba-小波多尺度建模方法

HCANet：基于分层Transformer架构的微表情识别模型

基于自适应池化注意力Transformer的唇语识别方法

注意力机制增强的输煤传送带异物检测