Non-invasive sensor based automated smoking activity detection

Abstract

Although smoking prevalence is declining in many countries, smoking related health problems still leads the preventable causes of death in the world. Several smoking intervention mechanisms have been introduced to help smoking cessation. However, these methods are inefficient since they lack in providing real time personalized intervention messages to the smoking addicted users. To address this challenge, the first step is to build an automated smoking behavior detection system. In this study, we propose an accelerometer sensor based non-invasive and automated framework for smoking behavior detection. We built a prototype device to collect data from several participants performing smoking and other five confounding activities. We used three different classifiers to compare activity detection performance using the extracted features from accelerometer data. Our evaluation demonstrates that the proposed approach is able to classify smoking activity among the confounding activities with high accuracy. The proposed system shows the potential for developing a real time automated smoking activity detection and intervention framework.

Zero-Shot Human Activity Recognition Using Non-Visual Sensors

“…Those sensors contribute to developing a wide range of application domains such as sport [38], human-computer interaction [39], surveillance [40], video streaming [41], healthcare system [42], and computer vision area [43]. Due to the properties of noninvasive sensors, some studies discussed how to monitor human activities using this type of sensors (i.e., non-visual sensors) because they are both easy to install and privacy preserving [44,45].…”

A New Approach for Smoking Event Detection Using a Variational Autoencoder and Neural Decision Forest

“…Given their good mobility and convenience, some research works have gradually started to use wristbands or smartwatches to detect smoking and other similar activities, such as drinking and eating. Maramis et al [23] and Bhandari et al [16] used only a single wrist gyroscope or accelerometer to detect cigarette-to-mouth gestures or smoking activities. Romeo [12] presented a smoking gesture detection system using a smartwatch with an accelerometer and a gyroscope sensor.…”

“…Regarding the learning techniques, previous works mostly use traditional classification methods such as the naï ve Bayes classifier [12,13], support vector machine [4,12,14], and random forest [12,[14][15][16][17][18]. Among these, the random forest classifier has proven its advantages and has been successfully used for smoking event detection and activity recognition in some works.…”

Ambulatory Smoking Habits Investigation based on Physiology and Context (ASSIST) using wearable sensors and mobile phones: protocol for an observational study

“…Besides, acoustic sensor and breath carbon monoxide (CO) sensor were also used as a way to identify smoking by Echebarria and Valencia, respectively 27 28. In addition, less cumbersome and more naturalistic systems with smaller form factor,s such as augmented lighters and a wrist-worn RisQ system, were used for capturing smoking events 29–32. These sensors can avoid users’ subjective bias and significantly alleviate their burden through automatic detection.…”