Enhanced Object Identification Through Deep Learning Techniques

Object identification involves identifying and locating objects in an image or video. This study aimed to build an advanced object identification system capable of accurately pinpointing and categorizing objects in visual data. This paper utilized annotated image datasets containing bounding boxes around the target object for model training. In this paper, we have used the YOLOv8 algorithm, through which we got better results. Confusion matrix, F1-confidence curve, recallconfidence curve, etc., were used to test the efficiency of the presented model. We have implemented the algorithm and have got about 0.98361 precision and 1 recall.