MCGS-SLAM (Multi-Camera Dataset)


IEEE International Conference on Robotics and Automation (ICRA), 2026


arXiv

Multi-Camera Airsim (MC-Airsim) Dataset

This AirSim Dataset is a multi-camera dataset designed for SLAM systems, collected using the AirSim simulator. We call it the MC-Airsim Dataset. We use some environments provided on Fab. The robot platform is a drone equipped with five cameras, named front_center, front_left, front_right, left_center, and right_center.

The dataset is organized as follows: 

├─images
│  ├─front_center
│  ├─front_left
│  ├─front_right
│  ├─left_center
│  └─right_center
└─poses
│  ├─front_center
│  ├─front_left
│  ├─front_right
│  └─left_center
│  └─right_center
└─camera_parameters.json

The images directory contains the image inputs from the five cameras. The poses directory stores the corresponding camera poses. The camera_parameters.json file provides the camera parameters, including both intrinsics and extrinsics.



# Env Ground Truth Image Pose Parameters
1 Factory GT1 Image Pose Camera Parameters
2 Village GT1 Image Pose Camera Parameters
3 Garden GT1 Image Pose Camera Parameters

Citation

If you find this dataset useful, please consider citing:

@article{cao2025mcgs,
  title={Mcgs-slam: A multi-camera slam framework using gaussian splatting for high-fidelity mapping},
  author={Cao, Zhihao and Wu, Hanyu and Tang, Li Wa and Luo, Zizhou and Zhang, Wei and Pollefeys, Marc and Zhu, Zihan and Oswald, Martin R},
  journal={arXiv preprint arXiv:2509.14191},
  year={2025}
}