Fusion of RetinaFace and improved FaceNet for individual cow identification in natural scenes

Cows’ posture change is the fatal influencing factor for accurate identification of individual cows. To achieve non-contact, high-precision detection and identification of individual cows in farm environment, a cow individual identification method by the fusion of RetinaFace and improved FaceNet was...

Full description

Saved in:
Bibliographic Details
Main Authors: Lingling Yang, Xingshi Xu, Jizheng Zhao, Huaibo Song
Format: Article
Language:English
Published: Elsevier 2024-12-01
Series:Information Processing in Agriculture
Subjects:
Online Access:http://www.sciencedirect.com/science/article/pii/S2214317323000653
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1850061853909057536
author Lingling Yang
Xingshi Xu
Jizheng Zhao
Huaibo Song
author_facet Lingling Yang
Xingshi Xu
Jizheng Zhao
Huaibo Song
author_sort Lingling Yang
collection DOAJ
description Cows’ posture change is the fatal influencing factor for accurate identification of individual cows. To achieve non-contact, high-precision detection and identification of individual cows in farm environment, a cow individual identification method by the fusion of RetinaFace and improved FaceNet was proposed. MobileNet-enhanced RetinaFace was applied to ameliorate the impact of output channel quantity and convolution kernel dynamics using depthwise convolution combined with pointwise convolution. Regression predictions of bovine facial features and keypoints were generated under varying distances, scales and sizes. FaceNet's core feature network was enhanced through MobileNet integration, and the loss function was jointly optimized with Cross Entropy Loss and Triplet Loss to achieve a quicker and more stable convergence curve. The distances between the generated embedding vectors of cow facial features were corresponding to the similarity between cow faces, enabling accurate matching. RetinaFace exhibited detection false negative rates of 2.67%, 0.66%, 2.67%, and 3.33% under conditions of occlusion, no occlusion, low light, and bright light for cow facial detection. For cow facial pattern detection, the false negative rates for black and white patterns, pure black and pure white were 1.33%, 6.00% and 8.00%, respectively. Regarding cow facial posture changes, the false negative rates for face upward, bowing down, profile, and normal posture were 1.33%, 1.33%, 4.00% and 0.66%, respectively. Improved FaceNet model achieved an accuray of 99.50% on training set and 83.60% on test set. In comparison to YOLOX, the recognition model presented in this research demonstrated increased accuracy in cow facial detection under occlusion, no occlusion and strong lighting conditions by 2.67%, 0.40%, and 0.40%, respectively. Moreover, the accuracy for patterns with pure black and pure white tones surpassed that of YOLOX by 1.06% and 5.71%, correspondingly. Additionally, the accuracy rates for face upward, bowing down, profile and normal posture were higher than YOLOX by 2.00%, 3.34%, 2.66% and 0.40%, respectively. The proposed model demonstrates the proficiency in accurately identifying individual cows in natural scenes.
format Article
id doaj-art-09fce6c2dc324bbbbc4aefedb5c76774
institution DOAJ
issn 2214-3173
language English
publishDate 2024-12-01
publisher Elsevier
record_format Article
series Information Processing in Agriculture
spelling doaj-art-09fce6c2dc324bbbbc4aefedb5c767742025-08-20T02:50:05ZengElsevierInformation Processing in Agriculture2214-31732024-12-0111451252310.1016/j.inpa.2023.09.001Fusion of RetinaFace and improved FaceNet for individual cow identification in natural scenesLingling Yang0Xingshi Xu1Jizheng Zhao2Huaibo Song3College of Mechanical and Electronic Engineering, Northwest A&F University, Yangling, Shaanxi 712100, China; Key Laboratory of Agricultural Internet of Things, Ministry of Agriculture and Rural Affairs, Yangling, Shaanxi 712100, China; Shaanxi Key Laboratory of Agricultural Information Perception and Intelligent Services, Yangling 712100, ChinaCollege of Mechanical and Electronic Engineering, Northwest A&F University, Yangling, Shaanxi 712100, China; Key Laboratory of Agricultural Internet of Things, Ministry of Agriculture and Rural Affairs, Yangling, Shaanxi 712100, China; Shaanxi Key Laboratory of Agricultural Information Perception and Intelligent Services, Yangling 712100, ChinaCollege of Mechanical and Electronic Engineering, Northwest A&F University, Yangling, Shaanxi 712100, China; Key Laboratory of Agricultural Internet of Things, Ministry of Agriculture and Rural Affairs, Yangling, Shaanxi 712100, China; Shaanxi Key Laboratory of Agricultural Information Perception and Intelligent Services, Yangling 712100, ChinaCorresponding author at: College of Mechanical and Electronic Engineering, Northwest A&F University, Yangling, Shaanxi 712100, China.; College of Mechanical and Electronic Engineering, Northwest A&F University, Yangling, Shaanxi 712100, China; Key Laboratory of Agricultural Internet of Things, Ministry of Agriculture and Rural Affairs, Yangling, Shaanxi 712100, China; Shaanxi Key Laboratory of Agricultural Information Perception and Intelligent Services, Yangling 712100, ChinaCows’ posture change is the fatal influencing factor for accurate identification of individual cows. To achieve non-contact, high-precision detection and identification of individual cows in farm environment, a cow individual identification method by the fusion of RetinaFace and improved FaceNet was proposed. MobileNet-enhanced RetinaFace was applied to ameliorate the impact of output channel quantity and convolution kernel dynamics using depthwise convolution combined with pointwise convolution. Regression predictions of bovine facial features and keypoints were generated under varying distances, scales and sizes. FaceNet's core feature network was enhanced through MobileNet integration, and the loss function was jointly optimized with Cross Entropy Loss and Triplet Loss to achieve a quicker and more stable convergence curve. The distances between the generated embedding vectors of cow facial features were corresponding to the similarity between cow faces, enabling accurate matching. RetinaFace exhibited detection false negative rates of 2.67%, 0.66%, 2.67%, and 3.33% under conditions of occlusion, no occlusion, low light, and bright light for cow facial detection. For cow facial pattern detection, the false negative rates for black and white patterns, pure black and pure white were 1.33%, 6.00% and 8.00%, respectively. Regarding cow facial posture changes, the false negative rates for face upward, bowing down, profile, and normal posture were 1.33%, 1.33%, 4.00% and 0.66%, respectively. Improved FaceNet model achieved an accuray of 99.50% on training set and 83.60% on test set. In comparison to YOLOX, the recognition model presented in this research demonstrated increased accuracy in cow facial detection under occlusion, no occlusion and strong lighting conditions by 2.67%, 0.40%, and 0.40%, respectively. Moreover, the accuracy for patterns with pure black and pure white tones surpassed that of YOLOX by 1.06% and 5.71%, correspondingly. Additionally, the accuracy rates for face upward, bowing down, profile and normal posture were higher than YOLOX by 2.00%, 3.34%, 2.66% and 0.40%, respectively. The proposed model demonstrates the proficiency in accurately identifying individual cows in natural scenes.http://www.sciencedirect.com/science/article/pii/S2214317323000653Cow face recognitionRetinaFaceFaceNetDeep learning
spellingShingle Lingling Yang
Xingshi Xu
Jizheng Zhao
Huaibo Song
Fusion of RetinaFace and improved FaceNet for individual cow identification in natural scenes
Information Processing in Agriculture
Cow face recognition
RetinaFace
FaceNet
Deep learning
title Fusion of RetinaFace and improved FaceNet for individual cow identification in natural scenes
title_full Fusion of RetinaFace and improved FaceNet for individual cow identification in natural scenes
title_fullStr Fusion of RetinaFace and improved FaceNet for individual cow identification in natural scenes
title_full_unstemmed Fusion of RetinaFace and improved FaceNet for individual cow identification in natural scenes
title_short Fusion of RetinaFace and improved FaceNet for individual cow identification in natural scenes
title_sort fusion of retinaface and improved facenet for individual cow identification in natural scenes
topic Cow face recognition
RetinaFace
FaceNet
Deep learning
url http://www.sciencedirect.com/science/article/pii/S2214317323000653
work_keys_str_mv AT linglingyang fusionofretinafaceandimprovedfacenetforindividualcowidentificationinnaturalscenes
AT xingshixu fusionofretinafaceandimprovedfacenetforindividualcowidentificationinnaturalscenes
AT jizhengzhao fusionofretinafaceandimprovedfacenetforindividualcowidentificationinnaturalscenes
AT huaibosong fusionofretinafaceandimprovedfacenetforindividualcowidentificationinnaturalscenes