Application of automated biometric technologies for human identification using anatomical or behavioural characteristics has multiplied manifold over the past decade, especially in large scale applications, both for security and social welfare initiatives. Soft computing is naturally amenable to biometric recognition for a variety of reasons. The feature extraction and the matching (classification) of biometrics is generally complex and non-linear that may be modelled better by soft computing rather than traditional models. While permanence is a required characteristic for biometrics, it is well known that biometric characteristics change with time and soft computing techniques provide the capability to model and be adaptive to these changes.
Biometric recognition is considered to be a human capability and soft computing techniques and generally inspired by and imitate biological function. Soft computing techniques are robust in situations commonly encountered in biometric recognition where traditional techniques fail – noisy, occluded, misaligned or deformed biometric data, leading to a high intra-class and a low inter-class variability.
Papers containing original and unpublished research are invited for the Special Session on Soft Computing Techniques for Biometric Technologies. The topics of interest, among others, include application of soft computing techniques for biometric sample quality estimation, sample enhancement, segmentation, ageing modelling, feature extraction, dimensionality reduction, biometric matching, multi-classifier or multi-biometric fusion, etc.