Kid on the phone! Toward automatic detection of children on mobile devices

Abstract

Studies have shown that children can be exposed to smart devices at a very early age. This has important implications on research in children-computer interaction, children online safety and early education. Many systems have been built based on such research. In this work, we present multiple techniques to automatically detect the presence of a child on a smart device, which could be used as the first step on such systems. Our methods distinguish children from adults based on behavioral differences while operating a touch-enabled modern computing device. Behavioral differences are extracted from data recorded by the touchscreen and built-in sensors.To evaluate the effectiveness of the proposed methods, a new data set has been created from 50 children and adults who interacted with off-the-shelf applications on smart phones. Results show that it is possible to achieve 99% accuracy and less than 0.5% error rate after 8 consecutive touch gestures using only touch information or 5 seconds of sensor reading. If information is used from multiple sensors, then only after 3 gestures, similar performance could be achieved.

Password-free Authentication for Smartphone Touchscreen Based on Finger Size Pattern

“…To determine the pattern range, the lowest and highest start positions of size 0.48923634 frequencies in the 5 data sets are taken. From Table8, the pattern-range is (10)(11)(12)(13)(14). = 1 passes the 0.7 threshold.…”

“…Ref. [11] is another study where they handled distinguishing between soft and hard finger pressure, based on users' features such as small fingers and speed. [12] proposes another powerful Android software development framework that uses OpenCV library for real-time face recognition.…”

A Survey of Privacy Vulnerabilities of Mobile Device Sensors

“…The authors exploited the k Nearest Neighbours (k-NN) algorithm, obtaining an accuracy of 85.3%. Similarly, Nguyen et al [115] developed a method to distinguish an adult from a child exploiting the behavioural differences captured by the motion sensors. Based on the hypothesis that children, with smaller hands, will tend to be more shaky, they achieved an accuracy of 96% using the Random Forest (RF) method.…”

“…The authors used touch interaction information to classify children into three groups aged 18 months to 8 years old. Nguyen et al [115] also conducted a study using RF on tap gestures to distinguish between an adult and a child, achieving an accuracy of 99%. Touchscreen data has also been used to extract a person's gender.…”

Protecting Mobile Food Diaries from Getting too Personal

“…Early studies [24,96] used high-dimensional features, and the main correlating feature for the specific attribute we study (gender) was the raw accelerometer trace. Moreover, in the recent past, many studies have been published regarding inferring demographic attributes such as gender and age [42,62,89] using raw accelerometer and gyroscope traces of wearables and mobile phones. In these papers, sensor data are again high-dimensional and raw, and need to be transformed in many ways to engineer features.…”