@article{CRSIJ26000180,
                  author = {Adeyemi, Biliqees Temitope, Makinde, Oladayo Ezekiel, Ojo, Olufemi Samuel},
                  title = {Improved Electric Eel Foraging Optimization-Based Convolutional Neural Network for Two-Level CAPTCHA Authentication Using Facial Emotion Recognition},
                  journal = {Cosmo Research and Science International Journal},
                  year = {2025},
                  volume = {1},
                  number = {6},
                  pages = {312-325},
                  issn = {3108-1584},
                  url = {https://cosmorsij.com/published/CRSIJ26000180.pdf},
                  abstract = {This study describes an Improved Electric Eel Foraging Optimization-Based Convolutional Neural Network (IEEFO-CNN) for two-level CAPTCHA validation based on facial expression recognition. The traditional CAPTCHAs are no longer resistant to the attacks of artificial intelligence, so the need for a more secure CAPTCHA to certify the cognitive and behavioural fact that the user is indeed a human still exists. An Improved Electric Eel Foraging Optimization algorithm incorporating adaptive energy factor control, elite preservation, dynamic migration, and Levy flight exploration was developed to optimize CNN hyper parameters. To put the proposed method to the test, we carried out an experimental evaluation using a real-world data set of 2,000 facial images from 500 different subjects. The numbers speak for themselves: our new IEEFO-CNN is shown to be an effective way of bolstering authentication security on the web and social media. With reference to performance, it has an accuracy of 98.50 per cent and an AUC of 0.98. False Acceptance Rate of 0.33%, a False Rejection Rate of 2.67% and an Equal Error Rate of 1.50% were also recorded.},
                  keywords = {Authentication, Algorithm, Hyper parameters, Security, Optimization, Improved },
                  month = {May}
        }