A Novel Deep Multifeature Extraction Framework Based on Attention Mechanism Using Wearable Sensor Data for Human Activity Recognition

Wang, Yan; Xu, Hongke; Wang, Mengmeng; Wang, Yuhao; Yang, Yang; Zhou, Shuang; Zeng, Jiaqi; Xu, Jie; Li, Shijie; Li, Jianjun

doi:10.1109/jsen.2023.3242603

Cited by 31 publications

(24 citation statements)

References 30 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Literature suggests that the UCI dataset typically results in lower accuracies, ranging from 0.8920 to 0.95, in comparison to the MotionSense dataset. For instance, several studies employing the UCI dataset, including [51] with a gated recurrent neural network, [52] with an inception network architecture, and [53] with a residual bi-directional LSTM, have observed similar performance trends. While these studies achieve results comparable to our findings, they concurrently underscore the inherent intricacies and limitations of the UCI dataset.…”

Section: ) Tl From Motionsense To Uci Datasetmentioning

confidence: 89%

TinyML Empowered Transfer Learning on the Edge

Hayajneh,

Hafeez,

Zaidi

et al. 2024

IEEE Open J. Commun. Soc.

View full text Add to dashboard Cite

Tiny machine learning (TinyML) is a promising approach to enable intelligent applications relying on Human Activity Recognition (HAR) on resource-limited and low-power internet of things (IoT) edge devices. However, designing efficient TinyML models for these devices remains challenging due to computational resource constraints and the need for customisation to unique use cases. To address this, we propose a novel approach that utilises transfer learning (TL) techniques on edge micro-controller units (MCUs) to accelerate TinyML development. Our strategy involves pre-training generalised ML models on large-scale and varied datasets and fine-tuning them on-device for specific applications using TL. We demonstrate the effectiveness of our approach for HAR by experimenting with a convolutional neural network, long short-term memory TL (CNN-LSTM-TL) model architecture and visualisation techniques like t-distributed stochastic neighbour embedding (t-SNE) for dimensionality reduction. To further validate our model's proficiency and adaptability, we conducted extensive testing using two distinct datasets: MotionSense and UCI. This dual-dataset approach allowed us to assess the robustness of our model across different data domains, showcasing its versatility and effectiveness in various HAR scenarios. Our results show significant model accuracy and reduced training time while maintaining high inference rates and low MCU memory footprint. We also provide insights into best practices for implementing TL on edge MCUs and evaluate classification performance metrics such as accuracy, precision, recall, F1 score, and categorical cross-entropy loss. Our work lays a solid foundation for faster and more efficient TinyML deployments through TL framework across different application domains and types of edge IoT devices.

show abstract

Section: ) Tl From Motionsense To Uci Datasetmentioning

confidence: 89%

TinyML Empowered Transfer Learning on the Edge

Hayajneh,

Hafeez,

Zaidi

et al. 2024

IEEE Open J. Commun. Soc.

View full text Add to dashboard Cite

show abstract

“…Wang et al [47] proposed a novel deep multi-feature extraction framework based on the attention mechanism (DMEFAM). Singh et al [22] utilized a self-attention mechanism to learn crucial time points for decoding human activity.…”

Section: Self-attention Mechanismmentioning

confidence: 99%

Enhancing Human Activity Recognition in Smart Homes with Self-Supervised Learning and Self-Attention

Chen,

Gouin-Vallerand,

Bouchard

et al. 2024

Sensors

View full text Add to dashboard Cite

Deep learning models have gained prominence in human activity recognition using ambient sensors, particularly for telemonitoring older adults’ daily activities in real-world scenarios. However, collecting large volumes of annotated sensor data presents a formidable challenge, given the time-consuming and costly nature of traditional manual annotation methods, especially for extensive projects. In response to this challenge, we propose a novel AttCLHAR model rooted in the self-supervised learning framework SimCLR and augmented with a self-attention mechanism. This model is designed for human activity recognition utilizing ambient sensor data, tailored explicitly for scenarios with limited or no annotations. AttCLHAR encompasses unsupervised pre-training and fine-tuning phases, sharing a common encoder module with two convolutional layers and a long short-term memory (LSTM) layer. The output is further connected to a self-attention layer, allowing the model to selectively focus on different input sequence segments. The incorporation of sharpness-aware minimization (SAM) aims to enhance model generalization by penalizing loss sharpness. The pre-training phase focuses on learning representative features from abundant unlabeled data, capturing both spatial and temporal dependencies in the sensor data. It facilitates the extraction of informative features for subsequent fine-tuning tasks. We extensively evaluated the AttCLHAR model using three CASAS smart home datasets (Aruba-1, Aruba-2, and Milan). We compared its performance against the SimCLR framework, SimCLR with SAM, and SimCLR with the self-attention layer. The experimental results demonstrate the superior performance of our approach, especially in semi-supervised and transfer learning scenarios. It outperforms existing models, marking a significant advancement in using self-supervised learning to extract valuable insights from unlabeled ambient sensor data in real-world environments.

show abstract

“…The development of smart devices has provided good opportunities for HAR based on wearable devices. Compared with computer vision-based recognition, HAR based on wearable devices has provided significant advantages such as low budget, high performance, and portability while avoiding the impact of video blind spots and insufficient illumination (Wang et al, 2023).…”

Section: Wearable Technological Products Which Used In Athleticsmentioning

confidence: 99%

“…With the appearance of wearable devices such as phones and tablets, the accessibility of temporal sensor data related to human activity has greatly increased. Therefore, the fields of human activity recognition (HAR), intelligent monitoring, human-computer interaction and video capturing based on wearable devices have also attracted more attention (Wang et al, 2023). In many sports, team sports, water sports, snow sports and of course individual sports such as running, cycling, triathlon, etc., wearable technological devices have become one of the safest ways to monitor the trainings (Chambers et al, 2015).…”

Section: Introductionmentioning

confidence: 99%

Innovative Developments in Basketball: Utilization of Technology

Kırıkoğlu

2023

The Use of Developing Technology in Sports

View full text Add to dashboard Cite

This research explores the transformative impact of the integration of technology in basketball, both on and off the court, from training and performance enhancement for athletes, coaches and conditioners, to data analytics for statistical coaches, administrators and spectator engagement. It also highlights the importance of the use of technology in player development, equipment development, injury prevention, pre-game and in-game strategies, and team-fan interaction, as well as the fact that technology not only improves the overall quality of basketball but also opens new avenues for growth and innovation in sport.

show abstract

A Novel Deep Multifeature Extraction Framework Based on Attention Mechanism Using Wearable Sensor Data for Human Activity Recognition

Cited by 31 publications

References 30 publications

TinyML Empowered Transfer Learning on the Edge

TinyML Empowered Transfer Learning on the Edge

Enhancing Human Activity Recognition in Smart Homes with Self-Supervised Learning and Self-Attention

Innovative Developments in Basketball: Utilization of Technology

Contact Info

Product

Resources

About