Attendance System using Eye or Iris Calibration

The Iris Attendance CASIA Project incorporates iris recognition technology into attendance systems in an advanced manner. With the CASIA Iris Dataset used for precise and reliable identification, this project starts structured with data preprocessing; thus, images are converted to grayscale and resized to 128x128 pixels, followed by normalization for consistency. In this case, with a pre-trained Iris segmentation model on Roboflow, iris regions are accurately detected and segmented, ensuring high-quality annotation. The processed and tagged data are used to train a CNN designed for the robust extraction of features, reduction of dimensionality, and classification. The Convolutional Neural Network (CNN) uses an Adam optimizer and sparse categorical cross-entropy loss during training to ensure efficient learning. Model evaluation metrics, like accuracy, validate the system’s performance, showing promise for precise iris recognition as well as its application in automated attendance systems.