Multi-Spectral Stereo (MS2) Dataset
Welcome to the Multi-Spectral Stereo Dataset!
Abstract
We introduce the multi-spectral stereo (MS2) outdoor dataset, including stereo RGB, stereo NIR, stereo thermal, stereo LiDAR data, and GPS/IMU information. Our dataset provides rectified and synchronized 184K data pairs taken from city, residential, road, campus, and suburban areas in the morning, daytime, and nighttime under clear-sky, cloudy, and rainy conditions. We designed the dataset to explore various computer vision algorithms from multi-spectral sensor data to achieve high-level performance, reliability, and robustness against challenging environments.
MS2 Dataset: examples
MS2 dataset provides:
(Synchronized) Stereo RGB images / Stereo NIR images / Stereo thermal images
(Synchronized) Stereo LiDAR scans / GPS/IMU navigation data
Projected depth map (in RGB, NIR, thermal image planes)
Odometry data (in RGB, NIR, thermal cameras, and LiDAR coordinates)
We visualized a few examples of MS2 dataset (Left image of each modality, depth map in thermal image plane)
Daytime + Clear-sky
Nighttime + Clear-sky
(multiple light sources)
Daytime + Rainy
Nighttime + Clear-sky
(few light sources)
Copyright
All datasets and benchmarks on this page are copyright by us and published under the Creative Commons Attribution-NonCommercial-ShareAlike 3.0 License. This means that you must attribute the work in the manner specified by the authors, you may not use this work for commercial purposes and if you alter, transform, or build upon this work, you may distribute the resulting work only under the same license.
Citation
When using the dataset in your research, please cite the following paper.
@inproceedings{shin2023deep,
title={Deep Depth Estimation From Thermal Image},
author={Shin, Ukcheol and Park, Jinsun and Kweon, In So},
booktitle={Proceedings of the IEEE/CVF Conference on Computer Vision and Pattern Recognition},
pages={1043--1053},
year={2023}
}
Related Pages
Dataset paper: [CVPR Paper], [Extended version]
Dataset loader & tool: [Github]
Monocular/Stereo depth benchmark: [Github]
Changelog
2023.05.31: The MS2 dataset is released.