• Corpus ID: 227126857

Exploring Contrastive Learning in Human Activity Recognition for Healthcare

@article{Tang2020ExploringCL,
  title={Exploring Contrastive Learning in Human Activity Recognition for Healthcare},
  author={Chi Ian Tang and Ignacio Perez-Pozuelo and Dimitris Spathis and Cecilia Mascolo},
  journal={ArXiv},
  year={2020},
  volume={abs/2011.11542},
  url={https://api.semanticscholar.org/CorpusID:227126857}
}
Preliminary results indicated an improvement over supervised and unsupervised learning methods when using fine-tuning and random rotation for augmentation, however, future work should explore under which conditions SimCLR is beneficial for HAR systems and other healthcare-related applications.
[PDF] Semantic Reader
99 Citations
Highly Influential Citations
23
Background Citations
51
Methods Citations
57
Results Citations
2

Figures from this paper

99 Citations

Sensor Data Augmentation by Resampling for Contrastive Learning in Human Activity Recognition

A new sensor data augmentation method by resampling is proposed, which simulates more realistic activity data by varying the sampling frequency to maximize the coverage of the sampling space, which outperforms all state-of-the-art methods under a small amount of labeled data.

RelCon: Relative Contrastive Learning for a Motion Foundation Model for Wearable Data

To the authors' knowledge, this work is the first to show the generalizability of a self-supervised learning model with motion data from wearables across distinct evaluation tasks.

Past, Present, and Future of Sensor-based Human Activity Recognition using Wearables: A Surveying Tutorial on a Still Challenging Task

This paper takes stock of sensor-based HAR -- surveying it from its beginnings to the current state of the field, and charting its future, accompanied by a hands-on tutorial, through which practitioners are guided in developing HAR systems for real-world application scenarios.

Using Self-supervised Learning Can Improve Model Fairness

It is demonstrated that SSL can significantly improve model fairness, while maintaining performance on par with supervised methods-exhibiting up to a 30% increase in fairness with minimal loss in performance through self-supervision.

A Washing Machine is All You Need? On the Feasibility of Machine Data for Self-Supervised Human Activity Recognition

This paper examines whether data collected by affixing sensors on running machinery, i.e., recording non-human movements/vibrations can also be utilized for self-supervised human activity recognition and observes that state-of-the-art methods perform surprisingly well relative to when utilizing large-scale human movement data.

Evaluating Fairness in Self-supervised and Supervised Models for Sequential Data

It is demonstrated that SSL has the capacity to achieve performance on par with supervised methods while significantly enhancing fairness--exhibiting up to a 27% increase in fairness with a mere 1% loss in performance through self-supervision.

Cross-Domain HAR: Few Shot Transfer Learning for Human Activity Recognition

A novel transfer learning framework–Cross-Domain HAR–which follows the teacher-student self-training paradigm to more effectively recognize activities with very limited label information is introduced, which enables substantial performance improvements over the state-of-the-art in sensor-based HAR scenarios.

Assessing the State of Self-Supervised Human Activity Recognition Using Wearables

This paper assesses the progress of self-supervised HAR research by introducing a framework that performs a multi-faceted exploration of model performance, and utilizes this framework to assess seven state-of-the-art self- supervised methods for HAR, leading to the formulation of insights into the properties of these techniques and to establish their value towards learning representations for diverse scenarios.

Study on temperature (τ ) variation for SimCLR-based activity recognition

A temperature (τ) variance study affecting the loss of SimCLR model and ultimately full HAR evaluation results is provided and regulation of the temperature(τ) coefficient is incorporated for improving the HAR feature qualities and overall performance for downstream tasks in healthcare setting.

Sensor Data Augmentation with Resampling for Contrastive Learning in Human Activity Recognition

The sampling frequency of sensors is investigated and the resampling augmentation method is proposed, which shows that the resampling augmentation method outperforms supervised learning by 9.88% on UCI HAR and 7.69% on Motion Sensor in the fine-tuning evaluation of contrast learning with a small amount of labeled data.
...

25 References

Deep, Convolutional, and Recurrent Models for Human Activity Recognition Using Wearables

This paper rigorously explore deep, convolutional, and recurrent approaches across three representative datasets that contain movement data captured with wearable sensors, and describes how to train recurrent approaches in this setting, introduces a novel regularisation approach, and illustrates how they outperform the state-of-the-art on a large benchmark dataset.

Multi-task Self-Supervised Learning for Human Activity Detection

A novel self-supervised technique for feature learning from sensory data that does not require access to any form of semantic labels, i.e., activity classes is proposed, which achieves performance levels superior to or comparable with fully-super supervised networks trained directly with activity labels, and it performs significantly better than unsupervised learning through autoencoders.

Exploring semi-supervised and active learning for activity recognition

Two different techniques to significantly reduce the required amount of labeled training data are explored and systematically analyzes based on semi-supervised learning and active learning and uses self-training and co-training.

Deep learning algorithms for human activity recognition using mobile and wearable sensor networks: State of the art and research challenges

UniMiB SHAR: A Dataset for Human Activity Recognition Using Acceleration Data from Smartphones

A new dataset of acceleration samples acquired with an Android smartphone designed for human activity recognition and fall detection is presented and shows that the presence of samples of the same subject both in the training and in the test datasets, increases the performance of the classifiers regardless of the feature vector used.

Human Activity Recognition for Healthcare using Smartphones

The experimental result on the publicly available dataset indicates that each of the sensors can be used for human activity recognition separately, however, accelerometer sensor data performed better than gyroscope sensor data with classification accuracy of 92%.

A Public Domain Dataset for Human Activity Recognition using Smartphones

An Activity Recognition database is described, built from the recordings of 30 subjects doing Activities of Daily Living while carrying a waist-mounted smartphone with embedded inertial sensors, which is released to public domain on a well-known on-line repository.

Activity Recognition Using Inertial Sensing for Healthcare, Wellbeing and Sports Applications: A Survey

This paper surveys the current research directions of activity recognition using inertial sensors, with potential application in healthcare, wellbeing and sports, and presents the main techniques utilized.

Smart Devices are Different: Assessing and MitigatingMobile Sensing Heterogeneities for Activity Recognition

It is indicated that on-device sensor and sensor handling heterogeneities impair HAR performances significantly and a novel clustering-based mitigation technique suitable for large-scale deployment of HAR is proposed, where heterogeneity of devices and their usage scenarios are intrinsic.

A Simple Framework for Contrastive Learning of Visual Representations

It is shown that composition of data augmentations plays a critical role in defining effective predictive tasks, and introducing a learnable nonlinear transformation between the representation and the contrastive loss substantially improves the quality of the learned representations, and contrastive learning benefits from larger batch sizes and more training steps compared to supervised learning.