Generative Adversarial Synthesis of Radar Point Cloud Scenes

TL;DR

This paper presents a GAN-based method for synthesizing realistic automotive radar point cloud scenes to aid in validation and verification, reducing the need for laborious real data collection.

Contribution

It introduces a novel radar scene synthesis approach using PointNet++ GANs, providing a scalable alternative to real dataset acquisition.

Findings

GAN-generated scenes achieve ~87% similarity to real scenes

The method offers a practical solution for radar data augmentation

Demonstrates effectiveness of PointNet++ GANs in radar scene synthesis

Abstract

For the validation and verification of automotive radars, datasets of realistic traffic scenarios are required, which, how ever, are laborious to acquire. In this paper, we introduce radar scene synthesis using GANs as an alternative to the real dataset acquisition and simulation-based approaches. We train a PointNet++ based GAN model to generate realistic radar point cloud scenes and use a binary classifier to evaluate the performance of scenes generated using this model against a test set of real scenes. We demonstrate that our GAN model achieves similar performance (~87%) to the real scenes test set.