Towards Self-Supervision for Video Identification of Individual Holstein-Friesian Cattle: The Cows2021 Dataset

In this paper we publish the largest identity-annotated Holstein-Friesian cattle dataset (Cows2021) and a first selfsupervision framework for video identification of individual animals. The dataset contains 10, 402 RGB images with labels for localisation and identity as well as 301 videos from the same herd. The data shows top-down in-barn imagery, which captures the breed’s individually distinctive black and white coat pattern. Motivated by the labelling burden involved in constructing visual cattle identification systems, we propose exploiting the temporal coat pattern appearance across videos as a self-supervision signal for animal identity learning. Using an individual-agnostic cattle detector that yields oriented bounding-boxes, rotationnormalised tracklets of individuals are formed via trackingby-detection and enriched via augmentations. This produces a ‘positive’ sample set per tracklet, which is paired against a ‘negative’ set sampled from random cattle of other videos. Frame-triplet contrastive learning is then employed to construct a metric latent space. The fitting of a Gaussian Mixture Model to this space yields a cattle identity classifier. Results show an accuracy of Top-1: 57.0% and Top-4: 76.9% and an Adjusted Rand Index: 0.53 compared to the ground truth. Whilst supervised training surpasses this benchmark by a large margin, we conclude that selfsupervision can nevertheless play a highly effective role in speeding up labelling efforts when initially constructing supervision information. We provide all data and full source code alongside an analysis and evaluation of the system.

[1]  Jing Gao,et al.  Visual Identification of Individual Holstein Friesian Cattle via Deep Metric Learning , 2020, ArXiv.

[2]  Ali Ismail Awad,et al.  From classical methods to animal biometrics: A review on cattle identification and tracking , 2016, Comput. Electron. Agric..

[3]  Sanjay Kumar Singh,et al.  Automatic identification of cattle using muzzle point pattern: a hybrid feature extraction and classification paradigm , 2017, Multimedia Tools and Applications.

[4]  Colin Greatwood,et al.  Visual Localisation and Individual Identification of Holstein Friesian Cattle via Deep Learning , 2017, 2017 IEEE International Conference on Computer Vision Workshops (ICCVW).

[5]  Jian Sun,et al.  Cow identification based on fusion of deep parts features , 2020 .

[6]  Salah Sukkarieh,et al.  Individual Cattle Identification Using a Deep Learning Based Framework , 2019, IFAC-PapersOnLine.

[7]  Jianguo Zhang,et al.  The PASCAL Visual Object Classes Challenge , 2006 .

[8]  William Andrew.  Visual biometric processes for collective identification of individual Friesian cattle , 2019 .

[9]  H. Kühl,et al.  Animal biometrics: quantifying and detecting phenotypic appearance. , 2013, Trends in ecology & evolution.

[10]  W. E. Petersen,et al.  The Identification of the Bovine by Means of Nose-Prints , 1922 .

[11]  Stephen I. Gallant,et al.  Perceptron-based learning algorithms , 1990, IEEE Trans. Neural Networks.

[12]  Douglas A. Reynolds,et al.  Gaussian Mixture Models , 2018, Encyclopedia of Biometrics.

[13]  William Andrew,et al.  Aerial Animal Biometrics: Individual Friesian Cattle Recovery and Visual Identification via an Autonomous UAV with Onboard Deep Inference , 2019, 2019 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS).

[14]  Hamdi A. Mahmoud,et al.  Bovines Muzzle Classification Based on Machine Learning Techniques , 2015 .

[15]  Neill W. Campbell,et al.  Automatic individual holstein friesian cattle identification via selective local coat pattern matching in RGB-D imagery , 2016, 2016 IEEE International Conference on Image Processing (ICIP).

[16]  Richard M. Everson,et al.  Video tracking of dairy cows for assessing mobility scores , 2013 .

[17]  Lin Wang,et al.  Automatic individual identification of Holstein dairy cows using tailhead images , 2017 .

[18]  Takashi Watanabe,et al.  Structural pattern recognition of biological textures with growing deformations: A case of cattle's muzzle patterns , 2004 .

[19]  Adrian R. Allen,et al.  Evaluation of retinal imaging technology for the biometric identification of bovine animals in Northern Ireland , 2008 .

[20]  Luciano Vieira Koenigkan,et al.  A Study on the Detection of Cattle in UAV Images Using Deep Learning , 2019, Sensors.

[21]  R Houston,et al.  A computerised database system for bovine traceability. , 2001, Revue scientifique et technique.

[22]  WangLin,et al.  Automatic individual identification of Holstein dairy cows using tailhead images , 2017 .

[23]  Georg Wendl,et al.  Electronic identification of cattle with injectable transponders , 1999 .

[24]  Alexander Mathis,et al.  Perspectives on Individual Animal Identification from Biology and Computer Vision , 2021, Integrative and comparative biology.

[25]  Majid Mirmehdi,et al.  Who Goes There? Exploiting Silhouettes and Wearable Signals for Subject Identification in Multi-Person Environments , 2019, 2019 IEEE/CVF International Conference on Computer Vision Workshop (ICCVW).

[26]  Ning Qian,et al.  On the momentum term in gradient descent learning algorithms , 1999, Neural Networks.

[27]  H. Robbins.  A Stochastic Approximation Method , 1951 .

[28]  Cassandra B Tucker,et al.  Branding Practices on Four Dairies in Kantale, Sri Lanka , 2018, Animals : an open access journal from MDPI.

[29]  Geoffrey E. Hinton,et al.  ImageNet classification with deep convolutional neural networks , 2012, Commun. ACM.

[30]  Gaël Varoquaux,et al.  Scikit-learn: Machine Learning in Python , 2011, J. Mach. Learn. Res..

[31]  Li Fei-Fei,et al.  ImageNet: A large-scale hierarchical image database , 2009, CVPR.

[32]  Jian Sun,et al.  Deep Residual Learning for Image Recognition , 2015, 2016 IEEE Conference on Computer Vision and Pattern Recognition (CVPR).

[33]  M. Cugmas,et al.  On comparing partitions , 2015 .

[34]  Ali Ismail Awad,et al.  Bag-of-Visual-Words for Cattle Identification from Muzzle Print Images , 2019, Applied Sciences.

[35]  Lucas Beyer,et al.  In Defense of the Triplet Loss for Person Re-Identification , 2017, ArXiv.

[36]  Kaiming He,et al.  Focal Loss for Dense Object Detection , 2017, 2017 IEEE International Conference on Computer Vision (ICCV).

[37]  Ross B. Girshick,et al.  Focal Loss for Dense Object Detection , 2017, IEEE Transactions on Pattern Analysis and Machine Intelligence.

[38]  Francis Butler,et al.  A framework for beef traceability from farm to slaughter using global standards: An Irish perspective , 2009 .

[39]  Luc Van Gool,et al.  The 2005 PASCAL Visual Object Classes Challenge , 2005, MLCW.

[40]  Aboul Ella Hassanien,et al.  Cattle Identification Using Muzzle Print Images Based on Texture Features Approach , 2014, IBICA.