B-HAR: an open-source baseline framework for in depth study of human activity recognition datasets and workflows

Abstract

Human Activity Recognition (HAR), based on machine and deep learning algorithms is considered one of the most promising technologies to monitor professional and daily life activities for different categories of people (e.g., athletes, elderly, kids, employers) in order to provide a variety of services related, for example to well-being, empowering of technical performances, prevention of risky situation, and educational purposes. However, the analysis of the effectiveness and the efficiency of HAR methodologies suffers from the lack of a standard workflow, which might represent the baseline for the estimation of the quality of the developed pattern recognition models. This makes the comparison among different approaches a challenging task. In addition, researchers can make mistakes that, when not detected, definitely affect the achieved results. To mitigate such issues, this paper proposes an open-source automatic and highly configurable framework, named B-HAR, for the definition, standardization, and development of a baseline framework in order to evaluate and compare HAR methodologies. It implements the most popular data processing methods for data preparation and the most commonly used machine and deep learning pattern recognition models.

Indoor Spatiotemporal Contact Analytics Using Landmark-Aided Pedestrian Dead Reckoning on Smartphones

“…The difference in the scale of each input variable increases the difficulty of the problem being modeled. If one of the features has a broad range of values, the objective functions of THE established model will be highly probably governed by the particular feature without normalization, suffering from poor performance during learning and sensitivity to input values and further resulting in a higher generalization error [ 35 ]. Therefore, the range of all data should be normalized so that each feature contributes approximately proportionately to the final result.…”

“…Therefore, the range of all data should be normalized so that each feature contributes approximately proportionately to the final result. Standardization makes the values of each feature in the data have zero means by subtracting from the mean in the numerator and unit variance, as shown in Equation (3): where the is the standardized data, represents the number of data channels, and and are the mean and standard deviations of the -th channel of the samples [ 35 ]. This method is widely used for normalization in many machine learning algorithms and is also adopted in this work to normalize the range of data we obtained.…”

A Systematic Evaluation of Recurrent Neural Network Models for Edge Intelligence and Human Activity Recognition Applications

“…Edge-based and adept HAR systems can replace expensive healthcare monitors if the solutions are accurate, reliable, low-powered, and small. However, the lacks of standard workflows and differences in evaluation protocols, evaluation metrics [24], data generation methods, and their quality make comparison of different approaches a challenging task and do not allow fair comparison of results [25,26].…”

“…This is an iterative procedure used to fetch light model solutions for an edge device. Reference [26,46] also used workflow designs similar to ours, but we used RNNs for training, and we did not evaluate power as a performance metric. In the context of HAR, or any other application, data collection and preprocessing are the most crucial components in the workflow and the most time consuming.…”

Human Activity Recognition Data Analysis: History, Evolutions, and New Trends

“…Demrozi [132], performs multiple experiments of many supervised techniques in many widely known Datasets such as WISDM, DAPHNET, PAPAM, HHAR (Phone), HHAR (watch), Mhealth, RSSI, CSI. For this, different algorithms are implemented such as KNN, LDA, QDA, RF, DT, CNN.…”