Symphony Lake Dataset consists of 121 visual surveys of a lakeshore over more than three years in Metz, France. Unique from roadway datasets, it adds breadth to a space at a time when larger and more diverse datasets are needed to train data hungry machine learning methods. Over 5 million images from an unmanned surface vehicle capture the unstructured, natural environment as it evolved over time. Significant variation in appearance is present on time scales of weeks, seasons, and years. Success in this space may demonstrate advancements in perception, SLAM, and environment monitoring.

Content

This is just a portion of the dataset covering the downsampled images recorded from the vehicle. They are organized by directories representing the month and datestamps and images containing the different positions and angles captured.

Acknowledgements

The full details and links to other components are available from the following original site:
http://dream.georgiatech-metz.fr/?q=node/79

Please cite the following paper if you use the dataset:

• Griffith, Shane; Chahine, Georges; Pradalier, Cédric; Symphony Lake Dataset, Submitted to IJRR, 2017.

Relevant Papers:

• Griffith, Shane; Pradalier, Cédric; Reprojection Flow for Image Registration Across Seasons, British Machine Vision Conference (BMVC), 2016

• Griffith, Shane; Pradalier, Cédric; Survey Registration for Long-Term Natural Environment Monitoring, Journal of Field Robotics, 2016