Impact of Occlusion Masks on Gender Classification from Iris Texture

Claudio Yáñez,Juan E. Tapia,Claudio A. Perez,Christoph Busch
DOI: https://doi.org/10.1049/2024/8526857
2024-01-27
IET Biometrics
Abstract:Gender classification on normalized iris images has been previously attempted with varying degrees of success. In these previous studies, it has been shown that occlusion masks may introduce gender information; occlusion masks are used in iris recognition to remove non-iris elements. When, the goal is to classify the gender using exclusively the iris texture, the presence of gender information in the masks may result in apparently higher accuracy, thereby not reflecting the actual gender information present in the iris. However, no measures have been taken to eliminate this information while preserving as much iris information as possible. We propose a novel method to assess the gender information present in the iris more accurately by eliminating gender information in the masks. This consists of pairing iris with similar masks and different gender, generating a paired mask using the OR operator, and applying this mask to the iris. Additionally, we manually fix iris segmentation errors to study their impact on the gender classification. Our results show that occlusion masks can account for 6.92% of the gender classification accuracy on average. Therefore, works aiming to perform gender classification using the iris texture from normalized iris images should eliminate this correlation.
computer science, artificial intelligence
What problem does this paper attempt to address?
### Problems the Paper Attempts to Solve This paper aims to address the issue of gender information being introduced by occlusion masks when using iris textures for gender classification. Specifically, the study proposes the following points: 1. **Eliminating Gender Information in Masks**: - A new method is proposed to eliminate gender information in masks, which involves pairing iris images with similar masks but different genders and using a logical OR operation to generate new paired masks. This method makes the mask distribution consistent for both genders, thereby eliminating gender information in the masks. 2. **Studying the Impact of Masks on Gender Classification**: - By manually correcting automatically generated masks, the study investigates the impact of these mask errors on gender classification results. 3. **Improving the Accuracy of Gender Classification Using Iris Textures**: - By eliminating gender information in the masks, the experimental results more accurately reflect the actual gender information present in the iris textures, rather than information from other non-iris regions. Overall, the core objective of this paper is to develop a method that can exclude additional gender information in masks when using iris textures for gender classification, thereby improving the accuracy and reliability of the classification.