RIE-SenseNet: Riemannian Manifold Embedding of Multi-Source Industrial Sensor Signals for Robust Pattern Recognition

TL;DR

RIE-SenseNet is a geometry-aware Transformer that embeds industrial sensor signals in a Riemannian manifold, improving pattern recognition robustness through hyperbolic sequence modeling and manifold-based augmentation.

Contribution

It introduces a novel Riemannian manifold embedding and a geometry-consistent augmentation technique for industrial sensor data analysis.

Findings

Achieves over 90% F1-score, outperforming CNN and Transformer baselines.

Effectively preserves sensor signal structure with non-Euclidean embeddings.

Demonstrates robustness in industrial pattern recognition tasks.

Abstract

Industrial sensor networks produce complex signals with nonlinear structure and shifting distributions. We propose RIE-SenseNet, a novel geometry-aware Transformer model that embeds sensor data in a Riemannian manifold to tackle these challenges. By leveraging hyperbolic geometry for sequence modeling and introducing a manifold-based augmentation technique, RIE-SenseNet preserves sensor signal structure and generates realistic synthetic samples. Experiments show RIE-SenseNet achieves >90% F1-score, far surpassing CNN and Transformer baselines. These results illustrate the benefit of combining non-Euclidean feature representations with geometry-consistent data augmentation for robust pattern recognition in industrial sensing.