Toward Optimal Sampling Rate Selection and Unbiased Classification for Precise Animal Activity Recognition

2026-04-01

AI Summary

This work addresses the challenge of insufficient recognition accuracy for minority behavior classes in animal behavior identification, which stems from suboptimal sampling rates and class imbalance. To this end, the authors propose the Individual Behavior-Aware Network (IBA-Net), which innovatively integrates multi-sampling-rate features with a classifier calibration mechanism. Specifically, IBA-Net employs a Mixture-of-Experts architecture to enable behavior-adaptive feature customization and incorporates an Equiangular Tight Frame (ETF) classifier grounded in neural collapse theory to achieve unbiased classification. Experimental results on three public datasets—covering goats, cattle, and horses—demonstrate that IBA-Net significantly outperforms existing methods, particularly in improving recognition accuracy for underrepresented behavior categories.

Abstract

With the rapid advancements in deep learning techniques, wearable sensor-aided animal activity recognition (AAR) has demonstrated promising performance, thereby improving livestock management efficiency as well as animal health and welfare monitoring. However, existing research often prioritizes overall performance, overlooking the fact that classification accuracies for specific animal behavioral categories may remain unsatisfactory. This issue typically stems from suboptimal sampling rates or class imbalance problems. To address these challenges and achieve high classification accuracy across all individual behaviors in farm animals, we propose a novel Individual-Behavior-Aware Network (IBA-Net). This network enhances the recognition of each specific behavior by simultaneously customizing features and calibrating the classifier. Specifically, considering that different behaviors require varying sampling rates to achieve optimal performance, we design a Mixture-of-Experts (MoE)-based Feature Customization (MFC) module. This module adaptively fuses data from multiple sampling rates, capturing customized features tailored to various animal behaviors. Additionally, to mitigate classifier bias toward majority classes caused by class imbalance, we develop a Neural Collapse-driven Classifier Calibration (NC3) module. This module introduces a fixed equiangular tight frame (ETF) classifier during the classification stage, maximizing the angles between pair-wise classifier vectors and thereby improving the classification performance for minority classes. To validate the effectiveness of IBA-Net, we conducted experiments on three public datasets covering goat, cattle, and horse activity recognition. The results demonstrate that our method consistently outperforms existing approaches across all datasets.

Problem

• animal activity recognition
• sampling rate selection
• class imbalance
• classification bias
• behavior-specific accuracy

Innovation

• Individual-Behavior-Aware Network
• Mixture-of-Experts
• Feature Customization
• Neural Collapse
• Classifier Calibration