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

TL;DR

This paper introduces IBA-Net, a novel deep learning framework that adaptively fuses multi-rate sensor data and calibrates classifiers to improve accuracy for all animal behaviors in activity recognition tasks.

Contribution

The paper proposes IBA-Net with MoE-based feature customization and ETF classifier calibration, addressing sampling rate optimization and class imbalance in animal activity recognition.

Findings

IBA-Net outperforms existing methods on goat, cattle, and horse datasets.

Adaptive sampling rate fusion improves behavior-specific feature extraction.

ETF classifier calibration enhances minority class recognition.

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.