Robust Activity Recognition via Redundancy-Aware CNNs and Novel Pooling for Noisy Mobile Sensor Data
Preprint
- 26 November 2025
- preprint
- Published by MDPI AG in Preprints.org
Abstract
This paper proposes a robust convolutional neural network (CNN) architecture for human activity recognition (HAR) using smartphone accelerometer data, evaluated on the WISDM dataset. We introduce two novel pooling mechanisms—Pooling A (Extrema Contrast Pooling (ECP)) and Pooling B (Center Minus Variation (CMV))—that enhance feature discrimination and noise robustness. Pooling A applies a nonlinear penalty based on the squared range between extrema, capturing sharp signal transitions, while Pooling B subtracts the standard deviation to penalize variability more smoothly, making it resilient to noise. To support these operations, input data are normalized to the [0,1] range, ensuring bounded, interpretable pooled outputs. Our dual-path framework includes: (1) a 1D CNN applied to raw tri-axial sensor streams with proposed pooling layers, and (2) a histogram-based image encoding pipeline that transforms segment-level sensor redundancy into RGB representations for a 2D CNN with fully connected layers. Across six activity classes, our system achieves up to 96.5% weighted classification accuracy, outperforming traditional average pooling and baseline models. Additionally, under Gaussian, salt-and-pepper, and mixed noise conditions, the proposed pooling layers consistently reduce performance degradation, demonstrating improved stability in real-world sensing environments. These results highlight the benefits of domain-informed pooling and data representations for mobile HAR systems, and offer a foundation for future IoT-based recommendation systems.Keywords
All Related Versions
- Version 1, 2025-11-26, Preprints
- Published version: Sensors, 26 (2), 710.
This publication has 11 references indexed in Scilit:
- WISNet: A deep neural network based human activity recognition systemExpert Systems with Applications, 2024
- Human Activity Recognition via Wi-Fi and Inertial Sensors With Machine LearningIEEE Access, 2024
- MarNASNets: Toward CNN Model Architectures Specific to Sensor-Based Human Activity RecognitionIEEE Sensors Journal, 2023
- Human activity recognition from multiple sensors data using deep CNNsMultimedia Tools and Applications, 2023
- Multi-level feature fusion for multimodal human activity recognition in Internet of Healthcare ThingsInformation Fusion, 2023
- Deep Learning in Human Activity Recognition with Wearable Sensors: A Review on AdvancesSensors, 2022
- Deep learning based human activity recognition (HAR) using wearable sensor dataInternational Journal of Information Management Data Insights, 2021
- WiFi CSI-Based Human Activity Recognition Using Deep Recurrent Neural NetworkIEEE Access, 2019
- A tutorial on human activity recognition using body-worn inertial sensorsACM Computing Surveys, 2014
- Activity recognition using cell phone accelerometersACM SIGKDD Explorations Newsletter, 2011