Comparação dos algoritmos YOLOv3 e SSD para identificação de objetos em imagens, em um dataset para navegação de robôs em ambientes interiores

Abstract

This work presents a study related to the research area of computer vision. The main objective is to compare YOLOv3 and SSD algorithms for identifying objects in images. In order to achieve the intended objective, at first, the algorithms were trained and compared without data augmentation. After, the data augmentation was executed for improving the performance of the algorithms. The YOLOv3 model presented a slightly better performance compared to the SSD, without performing data augmentation, because the F1-score for YOLOv3 is 75% and for SSD it is 73%. On the other hand, when performing data augmentation, the SSD model is favored, the F1-score obtained was 81% and, YOLOv3 obtained 80%.