Ethnicity- and Gender-based Subject Retrieval Using 3-D Face-Recognition Techniques

Abstract

While the retrieval of datasets from human subjects based on demographic characteristics such as gender or race is an ability with wide-ranging application, it remains poorly-studied. In contrast, a large body of work exists in the field of biometrics which has a different goal: the recognition of human subjects. Due to this disparity of interest, existing methods for retrieval based on demographic attributes tend to lag behind the more well-studied algorithms designed purely for face matching. The question this raises is whether a face recognition system could be leveraged to solve these other problems and, if so, how effective it could be. In the current work, we explore the limits of such a system for gender and ethnicity identification given (1) a ground truth of demographically-labeled, textureless 3-D models of human faces and (2) a state-of-the-art facerecognition algorithm. Once trained, our system is capable

A study on different experimental configurations for age, race, and gender estimation problems

“…Race estimation was also addressed in [60] where a LDA representation was developed under the Bayesian statistical decision framework: the task was formulated as a two-category classification problem to classify between Asian or non-Asian subjects. Finally, in [61], two of the above problems (the estimation of gender and race) were faced using facial imagery. In particular, the examination of the 3-D meshes of the face, without any associated texture or photographic information, was performed.…”

Large Scale 3D Morphable Models

“…Toderici et al (2010) perform an accurate estimation of gender and ethnicity based purely on the 3D facial shapes, without using any associated texture or photographic information. Their proposed method achieves around 99% accuracy for race and 94% for gender recognition.…”

3D Face Analysis: Advances and Perspectives

“…In 3D FER, it is desired to generate geometry features that are able to describe subject-independent properties of facial expressions [34]. In 3D GEC and FAE, gender, ethnicity, and age related clues are extracted to provide enough discrimination [35] [36] [37]. Meanwhile, the shape representation of 3D facial surfaces should also be robust to internal and external occlusions, particularly in 3D FR [38] [39].…”