Our paper presents FaceDetectCT, an innovative face recognition model designed to handle the complexities of diverse datasets and real-world environments. We provide a comprehensive overview of the model’s architecture, emphasizing its novel features. Extensive testing confirms that FaceDetectCT performs competitively across various metrics when compared to established benchmarks. While it shows promising accuracy, further refinements are necessary to enhance its adaptability to pose variations and improve computational efficiency. This research highlights our model’s potential to make significant strides in face recognition technology, with wide-ranging practical applications. Additionally, FaceDetectCT represents a pioneering integration of the Vision Transformer (ViT) architecture, specifically optimized to increase face detection resilience. Our extensive experimentation validates our model’s competitive performance across multiple metrics, including accuracy, precision, recall, F1 score, AUC-ROC, IoU, and computational efficiency. These metrics collectively demonstrate the model’s ability to perform reliably and efficiently in various scenarios, including those involving significant pose variations.