C2FHAR: Coarse-to-Fine Human Activity Recognition With Behavioral Context Modeling Using Smart Inertial Sensors

Today's smartphones and wearable devices come equipped with an array of inertial sensors, along with IMU-based Human Activity Recognition models to monitor everyday activities. However, such models rely on large amounts of annotated training data, which require considerable time and effort for collection. One has to recruit human subjects, define clear protocols for the subjects to follow, and manually annotate the collected data, along with the administrative work that goes into organizing such a recording.

Download Full-text

Adaptation and Selection Techniques Based on Deep Learning for Human Activity Recognition Using Inertial Sensors

Engineering Proceedings ◽

10.3390/ecsa-7-08159 ◽

2020 ◽

Vol 2 (1) ◽

pp. 22

Author(s):

Manuel Gil-Martín ◽

José Antúnez-Durango ◽

Rubén San-Segundo

Keyword(s):

Deep Learning ◽

Activity Recognition ◽

Human Activity ◽

Inertial Sensors ◽

Human Activity Recognition ◽

Training Data ◽

Generative Adversarial Networks ◽

Data Set ◽

Physical Activity Monitoring ◽

Training Data Selection

Deep learning techniques have been widely applied to Human Activity Recognition (HAR), but a specific challenge appears when HAR systems are trained and tested with different subjects. This paper describes and evaluates several techniques based on deep learning algorithms for adapting and selecting the training data used to generate a HAR system using accelerometer recordings. This paper proposes two alternatives: autoencoders and Generative Adversarial Networks (GANs). Both alternatives are based on deep neural networks including convolutional layers for feature extraction and fully-connected layers for classification. Fast Fourier Transform (FFT) is used as a transformation of acceleration data to provide an appropriate input data to the deep neural network. This study has used acceleration recordings from hand, chest and ankle sensors included in the Physical Activity Monitoring Data Set (PAMAP2) dataset. This is a public dataset including recordings from nine subjects while performing 12 activities such as walking, running, sitting, ascending stairs, or ironing. The evaluation has been performed using a Leave-One-Subject-Out (LOSO) cross-validation: all recordings from a subject are used as testing subset and recordings from the rest of the subjects are used as training subset. The obtained results suggest that strategies using autoencoders to adapt training data to testing data improve some users’ performance. Moreover, training data selection algorithms with autoencoders provide significant improvements. The GAN approach, using the generator or discriminator module, also provides improvement in selection experiments.

Download Full-text

Study on Temperature (τ) Variation for SimCLR based Activity Recognition

10.20944/preprints202107.0138.v2 ◽

2021 ◽

Author(s):

Pranjal Kumar

Keyword(s):

Activity Recognition ◽

Human Activity ◽

Human Activities ◽

Inertial Sensors ◽

Healthcare Setting ◽

Human Activity Recognition ◽

Stream Data ◽

Loss Mechanism ◽

Overall Performance ◽

Physiological Sensors

Human Activity Recognition (HAR) is a process to automatically detect human activities based on stream data generated from various sensors, including inertial sensors, physiological sensors, location sensors, cameras, time, and many others. Unsupervised contrastive learning has been excellent, while the contrastive loss mechanism is less studied. In this paper, we provide a temperature (τ) variance study affecting the loss of SimCLR model and ultimately full HAR evaluation results. We focus on understanding the implications of unsupervised contrastive loss in context of HAR data. In this work, also regulation of the temperature(τ) coefficient is incorporated for improving the HAR feature qualities and overall performance for downstream tasks in healthcare setting. Performance boost of 1.3% is observed in experimentation.

Download Full-text

Study on Temperature Variance for SimCLR based Activity Recognition

10.20944/preprints202107.0138.v1 ◽

2021 ◽

Author(s):

Pranjal Kumar

Keyword(s):

Activity Recognition ◽

Human Activity ◽

Human Activities ◽

Inertial Sensors ◽

Visual Representations ◽

Human Activity Recognition ◽

Stream Data ◽

Temperature Variance ◽

Learning Technique ◽

Physiological Sensors

Human Activity Recognition (HAR) is a process to automatically detect human activities based on stream data generated from various sensors, including inertial sensors, physiological sensors, location sensors, cameras, time, and many others. In this paper, we propose a robust SimCLR model for human activity recognition with a temperature variance study. In this work, SimCLR, a contrasting learning technique is optimized via regulating the temperature for visual representations, is incorporated for improving the HAR performance in healthcare.

Download Full-text

Human Activity Recognition Based on Residual Network and BiLSTM

Sensors ◽

10.3390/s22020635 ◽

2022 ◽

Vol 22 (2) ◽

pp. 635

Author(s):

Yong Li ◽

Luping Wang

Keyword(s):

Deep Learning ◽

Activity Recognition ◽

Human Activity ◽

Inertial Sensors ◽

Human Activity Recognition ◽

Activity Data ◽

Multidimensional Signals ◽

Proposed Model ◽

Public Datasets ◽

Residual Block

Due to the wide application of human activity recognition (HAR) in sports and health, a large number of HAR models based on deep learning have been proposed. However, many existing models ignore the effective extraction of spatial and temporal features of human activity data. This paper proposes a deep learning model based on residual block and bi-directional LSTM (BiLSTM). The model first extracts spatial features of multidimensional signals of MEMS inertial sensors automatically using the residual block, and then obtains the forward and backward dependencies of feature sequence using BiLSTM. Finally, the obtained features are fed into the Softmax layer to complete the human activity recognition. The optimal parameters of the model are obtained by experiments. A homemade dataset containing six common human activities of sitting, standing, walking, running, going upstairs and going downstairs is developed. The proposed model is evaluated on our dataset and two public datasets, WISDM and PAMAP2. The experimental results show that the proposed model achieves the accuracy of 96.95%, 97.32% and 97.15% on our dataset, WISDM and PAMAP2, respectively. Compared with some existing models, the proposed model has better performance and fewer parameters.

Download Full-text