DETECTION OF ROAD SURFACE DAMAGE USING CUSTOM OBJECTS THROUGH SINGLE SHOT DETECTION

Abstract

The aim of the research is to apply Single Shot Detection methods to identify road damage in custom objects. Despite the existence of various established techniques, the researchers are determined to continue using them because newer and improved approaches have become available. Additionally, it is important to delve into the underlying causes of road damage and introduce fresh ideas and reasoning into the detection process.  The research paper encompasses an exploration of techniques for data preparation, annotation of data labels, and the creation of a road damage dataset for training models.