International Journal of Computer Applications |
Foundation of Computer Science (FCS), NY, USA |
Volume 186 - Number 21 |
Year of Publication: 2024 |
Authors: Md Sayem Iftekar, Mohammed Aqib Zeeshan |
10.5120/ijca2024923507 |
Md Sayem Iftekar, Mohammed Aqib Zeeshan . The ‘face-api.js' Library for Accurate Face Recognition in Web- Applications and Possible use Cases with Accuracy Metrics. International Journal of Computer Applications. 186, 21 ( May 2024), 10-15. DOI=10.5120/ijca2024923507
This research paper explores the integration of role-based face login using the `face-api.js` framework, emphasizing its effectiveness in establishing robust face recognition-based login mechanisms and human sentiment detection. The study introduces a novel Face accuracy metrics formula to evaluate overall recognition correctness, addressing challenges in accurate facial feature extraction, real-time face detection and showed use cases where this library can be utilised in web application environment. Motivated by the need for secure authentication in web applications, the research employs pre-trained models for face detection, landmark identification, recognition, and sentiment analysis. The proposed methodology includes role-based user assignment, AI/ML algorithms, and countermeasures against spoofing attacks. The paper outlines the experimental results, demonstrating high accuracy in face and expression detection. The research contributes valuable insights into advancing secure authentication systems, paving the way for resilient and effective AI/ML-driven mechanisms in the digital landscape.