Generation of Training Data from HD Maps in the Lanelet2 Framework

TL;DR

This paper introduces lanelet2_ml_converter, an extension to the Lanelet2 HD map framework, enabling standardized training data generation and real-time map perception model input for automated driving systems.

Contribution

It provides a unified framework integrating HD map data with machine learning training and inference, filling a gap in standardized tools for map perception in automated driving.

Findings

Demonstrates usability of generated labels in machine learning tasks

Shows integration of map perception models with real-time data

Provides open-source implementation within Lanelet2 framework

Abstract

Using HD maps directly as training data for machine learning tasks has seen a massive surge in popularity and shown promising results, e.g. in the field of map perception. Despite that, a standardized HD map framework supporting all parts of map-based automated driving and training label generation from map data does not exist. Furthermore, feeding map perception models with map data as part of the input during real-time inference is not addressed by the research community. In order to fill this gap, we presentlanelet2_ml_converter, an integrated extension to the HD map framework Lanelet2, widely used in automated driving systems by academia and industry. With this addition Lanelet2 unifies map based automated driving, machine learning inference and training, all from a single source of map data and format. Requirements for a unified framework are analyzed and the implementation of these requirements is described. The usability of labels in state of the art machine learning is demonstrated with application examples from the field of map perception. The source code is available embedded in the Lanelet2 framework under https://github.com/fzi-forschungszentrum-informatik/Lanelet2/tree/feature_ml_converter