• Julian J

Human activity recognition (HAR) has a unique role in many important applications, including health-care services, ubiquitous computing, and smart buildings. Because smartphone sensors are non-intrusive, they are being used widely for identification of human activities. The signals received from smartphones are noisy, so feature engineering is performed on the data to extract more precise data. Various machine learning algorithms are used to recognize the different human activities. In this paper a comprehensive study is made on the five latest and modern technique for HAR. The techniques are Ensemble extreme learning machine, Mobile crowd sensing environment using optimized k -NN algorithm, Probabilistic First-Take-All, Convolution neural network model and Deep recurrent neural network. The major and common challenges that HAR is facing are discussed with the possible solutions. Some of the important applications for HAR are also discussed.

Convolution neural network, Deep recurrent neural network, Ensemble extreme learning machine, Human activity recognition, k -NN algorithm, Mobile crowdsensing, Probabilistic First-Take-All.

"HUMAN ACTIVITY RECOGNITION WITH SMARTPHONES", International Journal of Emerging Technologies and Innovative Research (www.jetir.org), ISSN:2349-5162, Vol.6, Issue 6, page no.196-202, June 2019, Available :http://www.jetir.org/papers/JETIR1906M25.pdf

"HUMAN ACTIVITY RECOGNITION WITH SMARTPHONES", International Journal of Emerging Technologies and Innovative Research (www.jetir.org | UGC and issn Approved), ISSN:2349-5162, Vol.6, Issue 6, page no. pp196-202, June 2019, Available at : http://www.jetir.org/papers/JETIR1906M25.pdf