Deep Learning‐Based Quantum System for Human Activity Recognition
Shoba Rani Salvadi, Pallati Narsimhulu, T. N. P. Madhuri · 2023
People's daily activities and communications with their living settings are becoming increasingly important to better comprehend through human activity recognition (HAR), a fiercely debated topic in ubiquitous computing environments. Social communication has always relied heavily on human behavior. In order to better understand human behavior, it is important to look at how people interact with each other. In a variety of applications, such as human-intelligent video surveillance, the identification of human behavior is a significant difficulty. Extraction and learning data are critical to the evaluation algorithm. Numerous imposing outcomes, including neural networks, came from the triumph of deep learning. In order to get superior outcomes, quantum computing is used in the deep learning model. ORQC-CNN (Optimized Random Quantum Circuits with Convolutional Neural Networks) model is used to identify the HAR. The architecture that consists of a series of quantum classified layer is shown as an analogy to the classical CNN. Artificial gorilla troops optimizer (AGTO) for ORQC-CNN parameter update is presented using variational quantum methods. According to a network complexity analysis, the proposed model outperforms its predecessor exponentially.