How Does Bilateral Ear Symmetry Affect Deep Ear Features?

2025-08-06

AI Summary

This study investigates the impact of bilateral ear asymmetry on deep earprint feature learning, demonstrating that neglecting left-right ear differences degrades model discriminability. To address this, we propose an explicit ear-side modeling approach: an ear-side classifier is integrated after the CNN backbone, and a side-separated training and testing paradigm is designed. Through cross-dataset evaluation, ablation studies, and alignment strategy optimization, we systematically assess the role of ear-side information. Our method achieves significant accuracy improvements across five public ear datasets, validating the effectiveness of ear-side–aware modeling for large-scale earprint recognition. This work constitutes the first systematic investigation revealing the critical role of bilateral ear asymmetry in deep earprint recognition and introduces a generalizable ear-side decoupling learning framework.

Technology Category

• Computer Vision: Other Foundations of Computer Vision
• Machine Learning: Multimodal Learning
• Intelligent Robots: Multimodal Perception & Sensor Fusion

Application Category

• Security and Privacy: Large-scale security measurements
• User Modeling, Personalization and Recommendation: Fairness-aware retrieval and ranking
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking

Abstract

Ear recognition has gained attention as a reliable biometric technique due to the distinctive characteristics of human ears. With the increasing availability of large-scale datasets, convolutional neural networks (CNNs) have been widely adopted to learn features directly from raw ear images, outperforming traditional hand-crafted methods. However, the effect of bilateral ear symmetry on the features learned by CNNs has received little attention in recent studies. In this paper, we investigate how bilateral ear symmetry influences the effectiveness of CNN-based ear recognition. To this end, we first develop an ear side classifier to automatically categorize ear images as either left or right. We then explore the impact of incorporating this side information during both training and test. Cross-dataset evaluations are conducted on five datasets. Our results suggest that treating left and right ears separately during training and testing can lead to notable performance improvements. Furthermore, our ablation studies on alignment strategies, input sizes, and various hyperparameter settings provide practical insights into training CNN-based ear recognition systems on large-scale datasets to achieve higher verification rates.

Problem

• Investigates bilateral ear symmetry impact on CNN features
• Develops ear side classifier for left-right categorization
• Explores performance gains from separate ear side training

Innovation

• Developed ear side classifier for left-right categorization
• Explored bilateral symmetry impact on CNN training
• Achieved higher verification rates via separate ear treatment