Multi-views Embedding for Cattle Re-identification

TL;DR

This paper explores cattle re-identification using deep CNNs, highlighting its unique challenges compared to human re-identification and evaluating various baseline methods through extensive ablation studies.

Contribution

It introduces a novel approach for cattle re-identification with comprehensive baseline comparisons and an in-depth ablation analysis of the task's unique challenges.

Findings

Deep CNNs can be adapted for cattle re-identification.

Cattle re-identification presents distinct challenges from human re-identification.

Baseline models show varying effectiveness, emphasizing the need for specialized solutions.

Abstract

People re-identification task has seen enormous improvements in the latest years, mainly due to the development of better image features extraction from deep Convolutional Neural Networks (CNN) and the availability of large datasets. However, little research has been conducted on animal identification and re-identification, even if this knowledge may be useful in a rich variety of different scenarios. Here, we tackle cattle re-identification exploiting deep CNN and show how this task is poorly related with the human one, presenting unique challenges that makes it far from being solved. We present various baselines, both based on deep architectures or on standard machine learning algorithms, and compared them with our solution. Finally, a rich ablation study has been conducted to further investigate the unique peculiarities of this task.