This document discusses various soft computing techniques for iris recognition, specifically focusing on two neural network approaches: Competitive neural network Learning Vector Quantization (LVQ) and Adaptive Resonance Associative Map (ARAM). It provides an overview of iris recognition as a biometric method, summarizes preprocessing steps like localization, segmentation, and normalization of iris images. It also describes feature extraction and matching steps. Finally, it defines artificial neural networks and discusses how LVQ and ARAM can be used for pattern matching in iris recognition applications.