Semi-Supervised Adversarial Learning Using LSTM for Human Activity Recognition

Abstract

The training of Human Activity Recognition (HAR) models requires a substantial amount of labeled data. Unfortunately, despite being trained on enormous datasets, most current models have poor performance rates when evaluated against anonymous data from new users. Furthermore, due to the limits and problems of working with human users, capturing adequate data for each new user is not feasible. This paper presents semi-supervised adversarial learning using the LSTM (Long-short term memory) approach for human activity recognition. This proposed method trains annotated and unannotated data (anonymous data) by adapting the semi-supervised learning paradigms on which adversarial learning capitalizes to improve the learning capabilities in dealing with errors that appear in the process. Moreover, it adapts to the change in human activity routine and new activities, i.e., it does not require prior understanding and historical information. Simultaneously, this method is designed as a temporal interactive model instantiation and shows the capacity to estimate heteroscedastic uncertainty owing to inherent data ambiguity. Our methodology also benefits from multiple parallel input sequential data predicting an output exploiting the synchronized LSTM. The proposed method proved to be the best state-of-the-art method with more than 98% accuracy in implementation utilizing the publicly available datasets collected from the smart home environment facilitated with heterogeneous sensors. This technique is a novel approach for high-level human activity recognition and is likely to be a broad application prospect for HAR.

Semi-Supervised Adversarial Auto-Encoder to Expedite Human Activity Recognition

“…Concurrent activity recognition, interleaved activity recognition, and the recognition average are the three sections that make up the analysis section. The general recognition accuracy is then contrasted with the other state-of-the-art approaches, including the CNN [ 44 ], the long short-term memory (LSTM) [ 45 ], and synchronized long short-term memory (Syn-LSTM) [ 46 ].…”

A New Approach for Human Activity Recognition (HAR) Using A Single Tri-axial Accelerometer Based on a Combination of Three Feature Subsets

“…Some examples of traditional machine learning were K-Nearest Neighbours (KNN) [18,25,26], Random Forest (RF) [26,27], Support Vector Machine (SVM) [18,26], Multi-Layer Perceptron [28], and AdaBoost [29]. While deep learning algorithms, such as CNN [19], LSTM [30], RNN [31], and others. Previous research showed [32,33], that some machine learning algorithms produced different performances, even with the same dataset.…”

Sensor-Based Human Activity and Behavior Research: Where Advanced Sensing and Recognition Technologies Meet

“…Despite the irreplaceable advantages of traditional feature-based machine learning suggested in Section 2.2 , deep learning is increasingly demonstrating its powerful adaptive capabilities. Besides [ 12 ], this Special Issue contains three more articles on deep learning [ 13 , 14 , 15 ], offering us multiple dimensions of thinking: The training of HAR models requires a large amount of annotated data corpus. Most current models are not robust when facing anonymized data from new users; meanwhile, capturing each new subject’s data is usually not possible.…”

“…Most current models are not robust when facing anonymized data from new users; meanwhile, capturing each new subject’s data is usually not possible. Yang et al described semi-supervised adversarial learning using the long-short term memory (LSTM) approach for HAR [ 13 ], which trains labeled and anonymous data by adapting the semi-supervised learning paradigms on which adversarial learning capitalizes to enhance the learning capabilities handling errors. The device-free, privacy-protected, and light-insensitive characteristics have pushed WIFI-based HAR technology into the limelight.…”