Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

MCD: Diverse Large-Scale Multi-Campus Dataset for Robot Perception

Thien-Minh Nguyen, Shenghai Yuan, Thien Hoang Nguyen, Pengyu Yin, Haozhi Cao, Lihua Xie, Maciej Wozniak, Patric Jensfelt, Marko Thiel, Justin Ziegenbein, Noel Blunder

TL;DR

A comprehensive dataset named MCD (Multi-Campus Dataset), featuring a wide range of sensing modalities, high-accuracy ground truth, and diverse challenging environments across three Eurasian university campuses, and introduces semantic annotations of 29 classes over 59k sparse NRE lidar scans across three domains.

Abstract

Perception plays a crucial role in various robot applications. However, existing well-annotated datasets are biased towards autonomous driving scenarios, while unlabelled SLAM datasets are quickly over-fitted, and often lack environment and domain variations. To expand the frontier of these fields, we introduce a comprehensive dataset named MCD (Multi-Campus Dataset), featuring a wide range of sensing modalities, high-accuracy ground truth, and diverse challenging environments across three Eurasian university campuses. MCD comprises both CCS (Classical Cylindrical Spinning) and NRE (Non-Repetitive Epicyclic) lidars, high-quality IMUs (Inertial Measurement Units), cameras, and UWB (Ultra-WideBand) sensors. Furthermore, in a pioneering effort, we introduce semantic annotations of 29 classes over 59k sparse NRE lidar scans across three domains, thus providing a novel challenge to existing semantic segmentation research upon this largely unexplored lidar modality. Finally, we propose, for the first time to the best of our knowledge, continuous-time ground truth based on optimization-based registration of lidar-inertial data on large survey-grade prior maps, which are also publicly released, each several times the size of existing ones. We conduct a rigorous evaluation of numerous state-of-the-art algorithms on MCD, report their performance, and highlight the challenges awaiting solutions from the research community.

MCD: Diverse Large-Scale Multi-Campus Dataset for Robot Perception

TL;DR

A comprehensive dataset named MCD (Multi-Campus Dataset), featuring a wide range of sensing modalities, high-accuracy ground truth, and diverse challenging environments across three Eurasian university campuses, and introduces semantic annotations of 29 classes over 59k sparse NRE lidar scans across three domains.

Abstract

Perception plays a crucial role in various robot applications. However, existing well-annotated datasets are biased towards autonomous driving scenarios, while unlabelled SLAM datasets are quickly over-fitted, and often lack environment and domain variations. To expand the frontier of these fields, we introduce a comprehensive dataset named MCD (Multi-Campus Dataset), featuring a wide range of sensing modalities, high-accuracy ground truth, and diverse challenging environments across three Eurasian university campuses. MCD comprises both CCS (Classical Cylindrical Spinning) and NRE (Non-Repetitive Epicyclic) lidars, high-quality IMUs (Inertial Measurement Units), cameras, and UWB (Ultra-WideBand) sensors. Furthermore, in a pioneering effort, we introduce semantic annotations of 29 classes over 59k sparse NRE lidar scans across three domains, thus providing a novel challenge to existing semantic segmentation research upon this largely unexplored lidar modality. Finally, we propose, for the first time to the best of our knowledge, continuous-time ground truth based on optimization-based registration of lidar-inertial data on large survey-grade prior maps, which are also publicly released, each several times the size of existing ones. We conduct a rigorous evaluation of numerous state-of-the-art algorithms on MCD, report their performance, and highlight the challenges awaiting solutions from the research community.
Paper Structure (13 sections, 2 equations, 8 figures, 3 tables)

This paper contains 13 sections, 2 equations, 8 figures, 3 tables.

Table of Contents

  1. Introduction
  2. Related Works
  3. Dataset Statistical Analysis
  4. Sequence Characteristics
  5. Semantic Annotations
  6. Survey Map Continuous-Time Registration
  7. Benchmarking on SOTA algorithms
  8. Lidar-Inertial SLAM
  9. Visual-Inertial SLAM
  10. Range-Aided Localization
  11. Semantic Segmentation
  12. Discussion and Insights
  13. Conclusion

Figures (8)

  • Figure 1: Up-to-scale comparision of survey-grade prior maps in MCD with other works.
  • Figure 2: Annotated NRE lidar scan motion concatenated over 20s period (left) and a single scan (right). This modality remains untested in semantic segmentation research to the best of our knowledge.
  • Figure 3: Example camera views at different campuses in MCD, with various terrains, backgrounds and lighting conditions. All faces are anonymized in compliance with local regulations.
  • Figure 4: The lidar point clouds with (left) and without (right) motion undistortion.
  • Figure 5: The sunglare and reflection outliers in the lidar scans are uncommon in lidar datasets. Significant effort were needed to segment out these outlier points.
  • ...and 3 more figures