Study on Facial Recognition Method Based on YOLOv5

Abstract

This study focuses on the facial recognition algorithm. To address limitations of facial recognition, this paper proposes a deep learning method called YOLOv5-Attention for facial feature extraction, recognition. The algorithm in the article is designed based on YOLOv5 and incorporates an attention mechanism module to enhance the expression ability of the model’s eyebrow features and improve its overall performance. The experimental results demonstrate that the algorithm proposed in this paper achieves a harmonic mean of 94% for eyebrow recognition accuracy. This highlights the effectiveness and reliability of the algorithm, which can be applied to eyebrow recognition and dramatic character facial recognition, as well as the research of general facial recognition.