LAUSR

Abstract Sports have become the important and most prominent play for each and every country to indulge their pride to the world. For this reason, countries are eager and interested in protecting the players/sportsmen in many ways such as health related, money, security etc. The life quality of sportsmen is improved by detecting huge potential known to be stress for preventing and managing the diseases. Moreover, low-cost wearable devices are available for monitoring the vital signs which leads to the detection of stress. Furthermore, the stress-levels are determined by using a particular vital sign known as Heart Rate Variability (HRV) that data is collected from a particular wearable device. In this paper, a real-time detection framework is proposed for analysing the level of stress for a particular sports person. The proposed framework consists of a hybrid classification technique named Multi-Output Regression (MOR) with Deep Convolutional Neural Networks (DCNN) to analyse and identify various stress levels and its relationship with data of HRV. Furthermore, 5-min time determination of each sportsman is distinguished based on their psychological and physical stress-levels. The simulation results show that the performance of the proposed framework obtains a high accuracy level when comparing with other models. With a lower error rate and based on efficiency, the proposed model achieves a high accuracy level of more than 96%.