Recognition of user activity with a combined image and accelerometer wearable sensor

Abstract

In emerging markets such as Vietnam, Human activity recognition is to identify the type of human activity from the received signal stream (image, accelerometer, etc.). This is a topic that has received a lot of attention from researchers and has opened up a new trend in robotics interaction in recent times. The activity recognition problem can be classified in two different ways depending on whether the type of sensor is used to collect activity signals or machine learning models for recognition. Authors use combination of methods: Experimental research and Test evaluation through survey. Authors also perform intensive research on previous related studies.The research of the topic is expected to be the foundation for feature selection orientation when building a recognition system through accelerometer, gyroscope and image sensors. And it is  expected to be the foundation for feature selection orientation when building recognition systems through accelerometer, gyroscope and image sensors.