labelCloud: A Lightweight Domain-Independent Labeling Tool for 3D Object\n Detection in Point Clouds

摘要

Within the past decade, the rise of applications based on artificial\nintelligence (AI) in general and machine learning (ML) in specific has led to\nmany significant contributions within different domains. The applications range\nfrom robotics over medical diagnoses up to autonomous driving. However, nearly\nall applications rely on trained data. In case this data consists of 3D images,\nit is of utmost importance that the labeling is as accurate as possible to\nensure high-quality outcomes of the ML models. Labeling in the 3D space is\nmostly manual work performed by expert workers, where they draw 3D bounding\nboxes around target objects the ML model should later automatically identify,\ne.g., pedestrians for autonomous driving or cancer cells within radiography.\n While a small range of recent 3D labeling tools exist, they all share three\nmajor shortcomings: (i) they are specified for autonomous driving applications,\n(ii) they lack convenience and comfort functions, and (iii) they have high\ndependencies and little flexibility in data format. Therefore, we propose a\nnovel labeling tool for 3D object detection in point clouds to address these\nshortcomings.\n