A Racing Dataset and Baseline Model for Track Detection in Autonomous Racing

TL;DR

This paper introduces RoRaTrack, a new multi-camera dataset for racing track detection, and proposes RaceGAN, a GAN-based baseline model that outperforms existing methods in challenging high-speed racing scenarios.

Contribution

The paper provides the first publicly available racing dataset with multi-camera annotations and introduces RaceGAN, a novel GAN-based model for robust track detection in racing environments.

Findings

RaceGAN outperforms current state-of-the-art models in track detection.

RoRaTrack effectively captures challenging racing conditions like blurriness and color inversion.

The dataset and code are publicly available for further research.

Abstract

A significant challenge in racing-related research is the lack of publicly available datasets containing raw images with corresponding annotations for the downstream task. In this paper, we introduce RoRaTrack, a novel dataset that contains annotated multi-camera image data from racing scenarios for track detection. The data is collected on a Dallara AV-21 at a racing circuit in Indiana, in collaboration with the Indy Autonomous Challenge (IAC). RoRaTrack addresses common problems such as blurriness due to high speed, color inversion from the camera, and absence of lane markings on the track. Consequently, we propose RaceGAN, a baseline model based on a Generative Adversarial Network (GAN) that effectively addresses these challenges. The proposed model demonstrates superior performance compared to current state-of-the-art machine learning models in track detection. The dataset and code for this work are available at https://github.com/ghosh64/RaceGAN.