BoviShift tracks cattle traits visually over six years and multiple animals
BoviShift: A Six-Year Multi-Attribute Audit of Visual Cattle Phenotyping Across Animals and Time
Computer Vision and Pattern Recognition
Summary
Estimating cattle traits like weight and breed using pictures and videos is tricky because many images can look alike or come from the same animals. The authors created BoviShift, a large collection of cattle images and videos gathered yearly over six years, to study how well computer vision can estimate traits across time and different animals. They found that how data is split for testing affects how well predictions work, and that some traits are easier to predict over time than others. Their study helps understand the challenges in tracking cattle traits visually with AI over multiple years.
What this means in practice
- •For livestock data teams: Evaluate and improve visual cattle trait estimation models using a six-year dataset capturing changes across animals and time.
- •For agriculture technology developers: Develop cattle monitoring tools that handle real-world changes in appearance and cohorts by training on multi-year annotated images and videos.
Authors
Mobasshir Bhuiya Shagor, Nilambar Halder Tonmoy, Md Saef Ullah Miah
Abstract
Computer vision offers a scalable, non-contact means of estimating livestock traits, but correlated images and video frames make reliable evaluation difficult. Image-random splitting can place views of the same animal in both training and test sets, while deployment further requires transfer to animals captured in future years. We introduce BoviShift, a corpus reconstructed from six consecutive annual cohorts of an operational cattle marketplace collected from 2021 to 2026. It contains 2{,}657 cattle, 10{,}610 still images, and 2{,}566 walk-around videos across nine recorded breed categories. We evaluate three splitting strategies through four configurations: image-random and pooled cattle-disjoint splits over the same development cohorts, followed by separate future-2025 and locked future-2026 evaluations. Using a fixed ResNet-50 multi-task baseline, we assess weight, age, height, breed, and coat color. Differences between image-random and cattle-disjoint splitting are task-dependent rather than a universal leakage penalty. Future-cohort behavior is also attribute- and cohort-specific: breed declines relative to pooled evaluation; mean weight and age performance fails to beat constant baselines in 2025; height remains above baseline; and color transfers well to 2025 but degrades in 2026. Multi-task training improves all five point estimates in a single-seed ablation, still-view aggregation helps selectively, and zero-shot video-frame aggregation exposes a still-to-video domain gap. To our knowledge, BoviShift is the longest-spanning cattle-vision resource combining consecutive annual cohorts with physical and visual attributes.