Forensic Iris

Forensic iris recognition uses iris patterns to identify deceased individuals, aiding forensic investigations. Current research focuses on improving the accuracy and interpretability of identification methods, particularly addressing challenges posed by post-mortem tissue decomposition. Convolutional neural networks, including architectures like VGG19 and ResNet, are employed for both iris recognition and estimating the post-mortem interval from iris images, while techniques incorporating human saliency improve model interpretability for forensic examiners. This emerging field holds significant potential for improving identification accuracy and providing valuable information for death investigations.