Cross-Species Transfer Learning in Agricultural AI: Evaluating ZebraPose Adaptation for Dairy Cattle Pose Estimation

TL;DR

This study evaluates the effectiveness of transfer learning from synthetic zebra images to real dairy cow pose estimation, highlighting domain gaps and the importance of diverse, realistic datasets for agricultural AI applications.

Contribution

It demonstrates the potential and limitations of cross-species transfer learning in livestock pose estimation, emphasizing the need for realistic datasets and robust models.

Findings

Combined model achieved AP=0.86 on in-distribution data

Significant generalization failures on unseen environments

Synthetic-to-real domain gap limits transfer effectiveness

Abstract

Pose estimation serves as a cornerstone of computer vision for understanding animal posture, behavior, and welfare. Yet, agricultural applications remain constrained by the scarcity of large, annotated datasets for livestock, especially dairy cattle. This study evaluates the potential and limitations of cross-species transfer learning by adapting ZebraPose - a vision transformer-based model trained on synthetic zebra imagery - for 27-keypoint detection in dairy cows under real barn conditions. Using three configurations - a custom on-farm dataset (375 images, Sussex, New Brunswick, Canada), a subset of the APT-36K benchmark dataset, and their combination, we systematically assessed model accuracy and generalization across environments. While the combined model achieved promising performance (AP = 0.86, AR = 0.87, PCK 0.5 = 0.869) on in-distribution data, substantial generalization failures occurred when applied to unseen barns and cow populations. These findings expose the synthetic-to-real domain gap as a major obstacle to agricultural AI deployment and emphasize that morphological similarity between species is insufficient for cross-domain transfer. The study provides practical insights into dataset diversity, environmental variability, and computational constraints that influence real-world deployment of livestock monitoring systems. We conclude with a call for agriculture-first AI design, prioritizing farm-level realism, cross-environment robustness, and open benchmark datasets to advance trustworthy and scalable animal-centric technologies.