# Aging-Invariant Sheep Face Recognition Through Feature Decoupling

## TL;DR

This paper introduces a new AI system for recognizing sheep faces that works accurately even as the sheep grow and change appearance.

## Contribution

The novel LBL-SheepNet framework decouples age-related and identity-specific features to maintain recognition accuracy across sheep growth stages.

## Key findings

- LBL-SheepNet achieved 95.5% identification accuracy in sheep face recognition.
- The framework uses adversarial learning to suppress age-biased features and focus on age-invariant identifiers.
- A dataset of 31,200 images from 55 sheep tracked monthly was constructed for training and evaluation.

## Abstract

To tackle the issue of maintaining accuracy in sheep face recognition across different growth stages, we constructed a dataset of 31,200 images from 55 sheep tracked monthly from 1 to 12 months of age. We proposed the LBL-SheepNet framework, which includes a Squeeze-and-Excitation (SE) module for enhancing feature representation, a nonlinear feature decoupling module for separating age-related features from identity-specific ones, and a correlation analysis module with adversarial learning to reduce age-biased interference. The framework achieved 95.5% identification accuracy and 95.3% average precision.

Precise recognition of individual ovine specimens plays a pivotal role in implementing smart agricultural platforms and optimizing herd management systems. With the development of deep learning technology, sheep face recognition provides an efficient and contactless solution for individual sheep identification. However, with the growth of sheep, their facial features keep changing, which poses challenges for existing sheep face recognition models to maintain accuracy across the dynamic changes in facial features over time, making it difficult to meet practical needs. To address this limitation, we propose the lifelong biometric learning of the sheep face network (LBL-SheepNet), a feature decoupling network designed for continuous adaptation to ovine facial changes, and constructed a dataset of 31,200 images from 55 sheep tracked monthly from 1 to 12 months of age. The LBL-SheepNet model addresses dynamic variations in facial features during sheep growth through a multi-module architectural framework. Firstly, a Squeeze-and-Excitation (SE) module enhances discriminative feature representation through adaptive channel-wise recalibration. Then, a nonlinear feature decoupling module employs a hybrid channel-batch attention mechanism to separate age-related features from identity-specific characteristics. Finally, a correlation analysis module utilizes adversarial learning to suppress age-biased feature interference, ensuring focus on age-invariant identifiers. Experimental results demonstrate that LBL-SheepNet achieves 95.5% identification accuracy and 95.3% average precision on the sheep face dataset. This study introduces a lifelong biometric learning (LBL) mechanism to mitigate recognition accuracy degradation caused by dynamic facial feature variations in growing sheep. By designing a feature decoupling network integrated with adversarial age-invariant learning, the proposed method addresses the performance limitations of existing models in long-term individual identification.