Automated Re-Identification of Holstein-Friesian Cattle in Dense Crowds

TL;DR

This paper introduces a new detect-segment-identify pipeline using open-vocabulary and segmentation models to improve re-identification of densely grouped Holstein-Friesian cattle, achieving high accuracy in farm CCTV data.

Contribution

It presents a novel pipeline combining segmentation and Re-ID networks with open-vocabulary models, significantly improving detection and re-identification in dense animal groups.

Findings

Achieved 98.93% detection accuracy in dense cattle groups.

Outperformed existing baselines by over 27% in accuracy.

Unsupervised contrastive learning yielded 94.82% Re-ID accuracy.

Abstract

Holstein-Friesian detection and re-identification (Re-ID) methods capture individuals well when targets are spatially separate. However, existing approaches, including YOLO-based species detection, break down when cows group closely together. This is particularly prevalent for species which have outline-breaking coat patterns. To boost both effectiveness and transferability in this setting, we propose a new detect-segment-identify pipeline that leverages the Open-Vocabulary Weight-free Localisation and the Segment Anything models as pre-processing stages alongside Re-ID networks. To evaluate our approach, we publish a collection of nine days CCTV data filmed on a working dairy farm. Our methodology overcomes detection breakdown in dense animal groupings, resulting in a 98.93% accuracy. This significantly outperforms current oriented bounding box-driven, as well as SAM species detection baselines with accuracy improvements of 47.52% and 27.13%, respectively. We show that unsupervised contrastive learning can build on this to yield 94.82% Re-ID accuracy on our test data. Our work demonstrates that Re-ID in crowded scenarios is both practical as well as reliable in working farm settings with no manual intervention. Code and dataset are provided for reproducibility.