Pit30M: A Benchmark for Global Localization in the Age of Self-Driving Cars

TL;DR

This paper introduces Pit30M, a large-scale, diverse dataset for evaluating retrieval-based localization methods in self-driving cars, along with benchmarking existing approaches and proposing a new LiDAR retrieval technique.

Contribution

The paper presents Pit30M, the largest dataset for city-scale localization, and provides a comprehensive benchmark for retrieval-based localization methods, including a novel LiDAR retrieval approach.

Findings

Pit30M dataset contains over 30 million frames with diverse conditions.

The proposed LiDAR retrieval method is competitive with state-of-the-art techniques.

Benchmark results establish a new standard for sub-metre localization accuracy.

Abstract

We are interested in understanding whether retrieval-based localization approaches are good enough in the context of self-driving vehicles. Towards this goal, we introduce Pit30M, a new image and LiDAR dataset with over 30 million frames, which is 10 to 100 times larger than those used in previous work. Pit30M is captured under diverse conditions (i.e., season, weather, time of the day, traffic), and provides accurate localization ground truth. We also automatically annotate our dataset with historical weather and astronomical data, as well as with image and LiDAR semantic segmentation as a proxy measure for occlusion. We benchmark multiple existing methods for image and LiDAR retrieval and, in the process, introduce a simple, yet effective convolutional network-based LiDAR retrieval method that is competitive with the state of the art. Our work provides, for the first time, a benchmark for sub-metre retrieval-based localization at city scale. The dataset, its Python SDK, as well as more information about the sensors, calibration, and metadata, are available on the project website: https://pit30m.github.io/