Face Recognition Using Optimized Feature Extraction and ML

This study presents a robust face recognition method combining HDG-HOG features.

This study proposes a robust face recognition framework addressing challenges like dataset variability and constrained environments. It combines HDG-HOG features with a Optimized Feature Vector (OFV), optimized using the Binary Grey Wolf Optimization (B-GWO) model. The methodology is validated on four public datasets, achieving high recognition accuracies: ORL (96.31%). Six machine learning models are used for classification, enabling accurate facial and expression recognition.