Comparative Analysis of Multi-pose Face Detection Between Yolo and Haar Cascade Classifier

Face detection is a biometric technology used to identify individuals based on their facial features. However, this technology faces challenges when detecting faces from various angles, which can affect the accuracy and speed of detection. To address this issue, algorithms such as You Only Look Once (YOLO) and the Haar Cascade Classifier have been used for object detection. YOLO is a real-time object detection algorithm, while the Haar Cascade Classifier is a simpler method that uses Haar-like features to detect objects. Several previous studies have tested both algorithms for object detection such as vehicles and crowd counting, with results showing that YOLO offers higher accuracy. This study aims to analyze the performance of YOLO and the Haar Cascade Classifier in detecting faces from various angles. The test results show that YOLO can consistently detect faces with 100% accuracy in all conditions. Meanwhile, the Haar Cascade Classifier also shows high accuracy, but experiences a significant drop at extreme angles of -90°. When the face is smiling in normal lighting, its accuracy is 36.96% for testing on images and 42.81% for testing on videos. Although the Haar Cascade Classifier has faster detection times, YOLO still excels in detection accuracy and consistency. Therefore, the algorithm selection can be tailored to the system's needs—whether prioritizing processing speed or detection accuracy.