Assessing the Influencing Factors on the Accuracy of Underage Facial Age Estimation
No ratings
Presented at CyberScience2020 2020 by
Swift response to the detection of endangered minors is an ongoing concern for law enforcement with therapid growth of disk capacities and data being stored in the cloud. Automated tools are needed to aid in CSEMinvestigation -- both to expedite the evidence discovery process, while lessening the investigator's exposure totraumatic material. In these investigations, age estimation techniques show great promise in helping decrease theoverflowing backlog of evidence obtained from the vast array of devices and online services. A lack of sufficienttraining data combined with natural human variance has been hindering accurate automated age estimation,especially for underage subjects. A comprehensive evaluation of the performance achieved on over 21,800underage subjects with two cloud age estimation services is presented, namely Amazon Web Service's Recognitionservice and Microsoft Azure's Face API. The objective of this work is to evaluate the influence that certain humanbiometric factors, facial expressions, and image quality, i.e., blur, noise, exposure and resolution, have on theoutcome of automated age estimation services. The thorough evaluation of the correlation and effects of such factorsaids the understanding of the performance and allows us to identify the most influencing factors to be overcome infuture age estimation modelling.