DiTer++

• DiTer++: Diverse Terrain and Multi-modal Dataset for Multi-Robot Navigation in Multi-session Outdoor Environments

  • Accepted in ICRA24 / Workshop on Future of Construction

  • Submitted to ICRA25

  

Download

• The dataset is provided on-demand. Link containing all data in our paper will be uploaded soon.

Sensor Configuration

• We configure multiple legged robot with various sensors offered in multi-session datasets.

  

Calibration

• Sensor calibration parameters including intrinsic & extrinsic will be uploaded soon.

Data & Topic

• Our data are stored in rosbag format
• Each agent has custom namespace of its own to be distinguishable
• Topic name for each sensor are listed in table below

  

Example Sequence in each sequence

• Our perceptual sensor data from each robot with heterogenous LiDAR.

  

• Multi-session sensor data is also avaliable from each site.

  

Map

• The global map is generated using Point-LIO, and Uni-Mapper.

Cite DiTer++

@article{kim2024diter,
  title={DiTer++: Diverse Terrain and Multi-modal Dataset for Multi-Robot Navigation in Multi-session Outdoor Environments},
  author={Kim, Juwon, Jeong, Seokhwan and Kim, Hogyun and Cho, Younggun}
}

Contact

• Juwon Kim ([email protected])
• Seokhwan Jeong ([email protected])
• Hogyun Kim ([email protected])