Abstract

Gender classification is an important topic in a wide variety of applications ranging from surveillance to selective marketing. Several recent studies have shown the predominance of local matching approaches in gender classifications results. Previous works in predicting gender-from-iris have relied on computing a separate set of textures representation. The state of the art shows that gender can be successfully predicted from the iris. There are clear computational advantages to predicting gender from the binary iris-code rather than computing another different texture representation. This topic brings new insights about the information present in the iris (and iris-code) to determine demographic information. The previous work adds evidence answering the fundamental question that the iris contains specific information about us, such as gender. The results, which show that gender classification from iris code is possible, will spur research to determine if other demographic factors (e.g., ethnicity, age, emotions) can also be predicted. This is an area of research that is overall in the early stages.