During this Covid-19 pandemic, the use of masks is highly recommended. The use of masks is very good to stop the spread of the virus. This is why many studies have created systems to assist humans in detecting the use of masks. However, from several existing studies, no research has been found that measures the accuracy of the detection system based on the direction of the object's head view, especially those using the fourth version of yolo (yoloV4). This study calculates accuracy based on 5 viewing angles, namely:(0, 45, 90) degrees sideways, up, and down. The distance between the camera and the test object is 1 meter. The results of the calculation, obtained very satisfactory results, namely 100% for each point of view. While the time required for the system to detect the fastest when the head position is straight ahead (0 degrees) with a time of 3.09 seconds. The longest time occurred when the head down position was 3.89 seconds