Object Detection, considered to be one of the basic fundamental and testing issues in Personal Computer vision, which is viewed as the extraordinary consideration in latest investigation. For the past 2 decades, it is been considered as an encapsulation of computer vision history. The objective of this paper is to look over the YOLOV5 and to evaluate the performance of YOLOV5 by various benchmarks and customized dataset. Motive of the study is to compare the performance of YOLOV5 by having the different data size, calculation speed and efficiency of detecting the objects, which are based on the dataset type. This work investigates and looks at the some of measurements for evaluating the YOLOV5 algorithm of detecting objects. Also audits the most utilized measurements for object detection and their disparities, applications, and primary ideas. It additionally proposes a standard execution that can be utilized as a benchmark among various datasets with least variation.