Journal of Cybersecurity and Information Management JCIM 2690-6775 2769-7851 10.54216/JCIM https://www.americaspg.com/journals/show/4161 2019 2019 Cascade Detection Technique for Face Mask ‎Recognition Based on YOLOv9 and CNN The College of Science for Women, University of Baghdad, Iraq Amal Amal The College of Science for Women, University of Baghdad, Iraq Wildan Jameel Hadi The College of Science for Women, University of Baghdad, Iraq Ammar Hussein Jassim Deep learning showed promise in many real-world applications. Recognition and item ‎identification are the most common. This publication tries to design and describe a system that can ‎classify people from images based on whether they are correctly wearing masks. The proposed ‎system is two-part. The first part is designed for facial detection using the YOLOv9 (You Only Look Once ‎version 9) compact deep learning model, which uses the mean intersection method over union to determine an optimal number of anchoring boxes and the Adam optimizer to improve facial detection efficacy. ‎The second component is a convolutional neural network for face feature extraction. These faces are ‎classified as a mask, without_mask, and incorrect_mask. These two components are integrated into ‎the proposed system for facemask recognition. ‎Empirical evaluations were conducted on the two self-collected datasets to train and evaluate the ‎proposed system's performance. The observed precision value of this system was 94.6% in the last ‎epoch; the recall value is 87.1%, and the mean average precision results are 92.74% as a face ‎detector. The classifier model train accuracy is 98.35%, and validation accuracy is 98.8%. Finally, the ‎comparative results indicated that the proposed framework was an effective model for face ‎detection, attaining a higher mean average precision value and outperforming other networks ‎assessed on the designated dataset for this task. The suggested network effectively detects and ‎classifies several faces in photos, including small faces in congested places. 2026 2026 103 119 10.54216/JCIM.170109 https://www.americaspg.com/articleinfo/2/show/4161