Guest Editorial Special Issue on Computational Intelligence for IoT-based Human Activity Recognition

Abstract

The papers in this special section focus on computational intelligence for Internet-of-Things-based human activity recognition (HAR). HAR benefits numerous real-world applications. For instance, it can be adopted in a healthcare service system to monitor the rehabilitation processes of patients. Another important application of HAR is in security and surveillance which require to analyze human behaviors and detect anomalies in specific areas. Finally, in human computer interface, recognizing human activities can be used to control robots and play virtual reality games. Many Internet of Things (IoT) sensors have been utilized for human activity recognition, such as wearable sensors, smartphones, radio frequency (RF) sensors. Owing to the rapid development of wireless sensor networks in IoT, a large amount of data have been collected for the recognition of human activities with different types of sensors. Conventional computational intelligence algorithms, such as shallow neural networks, require to manually extract some representative features from large and noisy sensory data, which may hinder their performance in real-world applications. Alternatively, the more advanced computational intelligence algorithms of deep neural networks have achieved great successes in many challenging research areas, such as image recognition and natural language processing. The key merit of the deep neural networks is the ability to automatically learn more accurate representative features from massive amount of data, without going through the manual and time-consuming feature extraction process.