Upon reading a couple of postings on Improbable Research (both postings, below), my first question is: Why would anyone care to do this? That is, what is the importance of the capability?
Computerised gender classification (part 1: eyebrows)
September 11th, 2015
Can you tell a person’s gender from just a glimpse of their eyebrows? Could a computerised system do the same? To find out, a project was undertaken by Yujie Dong (of the Holcombe Department of Electrical and Computer Engineering, Clemson University) and Damon L. Woodard (now at the Biometrics and Pattern Recognition Lab, University of Florida) in 2011. The team used Minimum Distance Classifier (MD), Linear Discriminant Analysis Classifier (LDA) and Support Vector Machine Classifier (SVM) algorithms to extract eight eyebrow features from images (rectangularity, eccentricity, and isoperimetric quotient etc.) and were then able to determine a subject’s gender with an average accuracy of 96%
Because of their comparatively robust demarcation, eyebrows, say the team, provide a more reliable method of gender classification when compared to other ‘periocular’ [around the eye] features such as skin wrinkles – especially when the images are non-ideal, or, as some describe them ‘in-the-wild’. See: ‘Eyebrow shape-based features for biometric recognition and gender classification: A feasibility study’ (in: Biometrics (IJCB), 2011 International Joint Conference on. IEEE, 2011, pp. 1–8.)
Coming soon : Computerised gender classification (part 2)
Computerised gender classification (part 2, shoes etc)
September 12th, 2015
The previous article in this series drew attention to the possibilities offered by eyebrows in the field of automated computerised gender classification. But eyebrows are by no means the only choice available to classifiers. For a recent (2015) roundup of the options on offer, see the work of Yingxiao Wu, Yan Zhuang, Xi Long, and Feng Lin, ‘Human Gender Classification: A Review’ (in: IEEE Sensors Journal, epub 2015)
The team first outline the importance of the field:
“The aim of gender classification is to recognize the gender of a person based on the characteristics differentiating between masculinity and femininity. It is considered as the one of the most important application [sic] using pattern recognition method in the area of artificial intelligence.”
And then go on to list the currently documented possibilities. [See diag. above] They include: body, eyebrow, handshape, fingernail, gait, gesture, hairstyle, footwear, voice, blogging style etc etc.
The reasons for caring about the capability, as given in the Human Gender Classification: A Review, are below.
To be concrete, the applications of a gender classification technique can be utilized in the following fields.
A. Human-Computer Interaction In the field of HCI [14], the robots or computers need to identify and verify the human gender to improve the system performance based on the personalized information. By successful determining the gender, the system can provide more appropriate and customized service for the customer by adapting to users according to their genders [5].
B. Surveillance System Classifying gender in surveillance system for public places (e.g., airport, train station, shopping mall) can serve to enhance the security such as counting the number of males and females, whose aim is to restrict access of people entering to certain areas. Another application is to assist the intelligent security and surveillance system to track moving object and detect abnormal behaviors. In addition, this can help better evaluate the threat level for a specific gender if the gender information can be obtained in advance [13].
C. Commercial Development Gender classification is useful for guiding effective marketing and establishing smart shopping environment, in which the production can be directed to specific users through websites, electronic marketing, advertising [15], etc. For instance, in a supermarket or department store, knowing the number of male and female customers helps the store managers to make efficient sales and managing decisions.
D. Demographic Research The application of a gender classification system helps the demographic research in efficiently collecting demographic information [7]. Automatic identification of human gender enhances the demographic statistics (e.g. gender, disability status, race definition) and population prediction [16]. The ability to automatically detect gender information acts as a supplementary method for existing demographic research on the web or in public places [17].
I am flummoxed. I can't recall who said it; but, about 40 years ago someone wrote to the effect that unless one wishes to have sex with another person, why should it matter what sex is the other person.
Recent Comments