UAC: Uncertainty-Aware Calibration of Neural Networks for Gesture Detection

2025-04-02

AI Summary

IMU-based gesture recognition for safety-critical applications (e.g., construction, healthcare) suffers from poor probability calibration and low robustness to out-of-distribution (OOD) data. To address these limitations, we propose the first two-stage uncertainty-aware calibration framework: (1) an uncertainty modeling neural network jointly predicts class labels and prediction entropy; (2) an entropy-weighted multi-window expectation mechanism dynamically calibrates output probabilities. Evaluated on three public IMU gesture datasets, our method significantly outperforms state-of-the-art calibration approaches—including temperature scaling, entropy maximization, and Laplace approximation—reducing expected calibration error (ECE) by over 37% while simultaneously improving classification accuracy. Notably, it is the only method demonstrated to enhance OOD calibration performance. This work establishes a new paradigm for high-reliability, uncertainty-aware AI in human–machine interaction.

Technology Category

• Machine Learning: Calibration & Uncertainty Quantification
• Reasoning under Uncertainty: Uncertainty Representations
• Intelligent Robots: State Estimation

Application Category

• Responsible Web: Machine-in-the-loop, human agency and autonomy
• User Modeling, Personalization and Recommendation: On-Device user modeling, personalization, and recommendation
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking

Abstract

Artificial intelligence has the potential to impact safety and efficiency in safety-critical domains such as construction, manufacturing, and healthcare. For example, using sensor data from wearable devices, such as inertial measurement units (IMUs), human gestures can be detected while maintaining privacy, thereby ensuring that safety protocols are followed. However, strict safety requirements in these domains have limited the adoption of AI, since accurate calibration of predicted probabilities and robustness against out-of-distribution (OOD) data is necessary. This paper proposes UAC (Uncertainty-Aware Calibration), a novel two-step method to address these challenges in IMU-based gesture recognition. First, we present an uncertainty-aware gesture network architecture that predicts both gesture probabilities and their associated uncertainties from IMU data. This uncertainty is then used to calibrate the probabilities of each potential gesture. Second, an entropy-weighted expectation of predictions over multiple IMU data windows is used to improve accuracy while maintaining correct calibration. Our method is evaluated using three publicly available IMU datasets for gesture detection and is compared to three state-of-the-art calibration methods for neural networks: temperature scaling, entropy maximization, and Laplace approximation. UAC outperforms existing methods, achieving improved accuracy and calibration in both OOD and in-distribution scenarios. Moreover, we find that, unlike our method, none of the state-of-the-art methods significantly improve the calibration of IMU-based gesture recognition models. In conclusion, our work highlights the advantages of uncertainty-aware calibration of neural networks, demonstrating improvements in both calibration and accuracy for gesture detection using IMU data.

Problem

• Calibrating neural networks for accurate gesture probability prediction
• Improving robustness against out-of-distribution IMU data
• Enhancing accuracy and calibration in gesture detection

Innovation

• Uncertainty-aware gesture network architecture
• Entropy-weighted expectation of predictions
• Improved calibration and accuracy in gesture detection