Evaluation shows ResNet-50 achieves 99.60% accuracy in face recognition, highlighting improvements in pose variations and feature extraction.
Face recognition remains a critical task in computer vision, especially in applications involving authentication, surveillance, and access control. However, real-world challenges like pose variations and occlusions continue to hinder recognition performance. This paper presents a pose-inclusive face recognition system that integrates Multi-task Cascaded Convolutional Networks (MTCNN) for face detection and alignment with ResNet-50 for feature extraction and classification. The Labeled Faces in the Wild (LFW) dataset is used, and images undergo data augmentation and normalization to simulate pose diversity. Experimental evaluation demonstrates that ResNet-50 significantly outperforms traditional CNN models, achieving an accuracy of 99.60%. The proposed approach ensures robust and scalable performance in uncontrolled environments.
No takes yet. Share an insight, caveat, or question.
Harshitha et al. (2025) studied this question.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: