Document Type : Original Article

Authors

1 Department of Computer Engineering, Faculty of Technology and Engineering- East of Guilan, University of Guilan, Guilan, Iran.

2 Department of Computer Engineering, Faculty of Technology and Engineering- East of Guilan, University of Guilan, Guilan, Iran

3 Department of Computer Science, University of Applied Sciences Bonn-Rhein-Sieg, Germany

4 Faculty of Technology and Engineering, University of Guilan

Abstract

Accurate identification of individual dogs plays a crucial role in various applications including pet recovery, veterinary management, and animal welfare. This study proposes a fully automated dog identification framework based on unique nose-print biometric patterns, leveraging deep learning techniques to overcome limitations of traditional identification methods. The proposed approach processes user-submitted videos by selecting the most frontal frame via head pose estimation, detects the nose region using a fine-tuned YOLOv8 model, and extracts discriminative embeddings through a multi-resolution ResNeSt-based convolutional network enhanced with advanced augmentation strategies. The resulting embeddings are fused to produce robust identity descriptors capable of distinguishing between thousands of individual dogs. Experimental results demonstrate the system’s efficacy under real-world conditions, emphasizing its potential for practical deployment in pet identification and management systems.

Keywords

[1] K. Lai, X. Tu and S. Yanushkevich, (2019) "Dog Identification using Soft Biometrics and Neural Networks," 2019 International Joint Conference on Neural Networks (IJCNN), Budapest, Hungary, pp. 1–8, doi: 10.1109/IJCNN.2019.8851971.
[2] T. Moreira, M. Perez, R. Werneck, and E. Valle, (2017) "Where is my puppy? Retrieving lost dogs by facial features," Multimedia Tools and Applications, vol. 76, doi: 10.1007/s11042-016-3824-1.
[3] Y. Qiao, D. Su, H. Kong, S. Sukkarieh, S. Lomax, and C. Clark, (2019) "Individual cattle identification using a deep learning based framework," IFAC-PapersOnLine, vol. 52, no. 30, pp. 318–323, doi: 10.1016/j.ifacol.2019.12.558.
[4] E. Azizi and L. Zaman, (2023) "Deep learning pet identification using face and body," Information, vol. 14, no. 5, p. 278, doi: 10.3390/info14050278.
[5] S. Schneider, G. W. Taylor, and S. C. Kremer, (2020) "Similarity learning networks for animal individual re-identification - beyond the capabilities of a human observer," in 2020 IEEE Winter Applications of Computer Vision Workshops (WACVW), Snowmass, CO, USA, 2020, pp. 44-52, doi: 10.1109/WACVW50321.2020.9096925.
[6] A. Voinea, R. Kock, and M. A. Dhali, (2025) "LostPaw: Finding lost pets using a contrastive learning-based transformer with visual input," in Proc. 14th Int. Conf. Pattern Recognit. Appl. Methods (ICPRAM), pp. 757–763, doi: 10.5220/0013261600003905.
[7] Alibaba Cloud Tianchi, "CVPR2022 Biometrics Workshop – Pet Biometric Challenge," [Online]. Available: https://tianchi.aliyun.com/competition/entrance/531952/information
[8] Ultralytics, "YOLOv8 models," Ultralytics Documentation, 2024. [Online]. Available: https://docs.ultralytics.com/models/.
[9] H. Zhang, et al., "ResNeSt: Split-Attention Networks," GitHub repository, [Online]. Available: https://github.com/zhanghang1989/ResNeSt/releases/download/weights_step1/resnest101-22405ba7.pth
[10] Fei Shen, Zhe Wang, Zijun Wang, Xiaode Fu, Jiayi Chen, Xiaoyu Du, and Jinhui Tang, “A Competitive Method for Dog Nose-print Re-identification,” Nanjing University of Science and Technology, DeepBlue Technology Co., Ltd, Guangdong University of Technology.
[11] B. Li, Z. Wang, N. Wu, S. Shi, and Q. Ma, "Dog nose print matching with dual global descriptor based on contrastive learning," Yunnan University, ShanghaiTech University, and Zhejiang A&F University.
[12] Z. Li, Z. Li, H. Zheng, B. Liu, R. Wang, Y. Shi, J. Chen, and H. Ling, "1st place solution to Pet Biometric Challenge 2022," Huazhong University of Science and Technology.