๐—ฃ๐—ฎ๐˜ƒ๐—ฒ๐—ฑ ๐—ผ๐—ฟ ๐˜‚๐—ป๐—ฝ๐—ฎ๐˜ƒ๐—ฒ๐—ฑ? ๐Ÿ›ฃ๏ธ
Accurate road surface information is crucial for emergency responses and route planning. But #OpenStreetMap has comprehensive surface attributes for only 30โ€“40% of the roads worldwide.

Using state-of-the-art GeoAI methods, we created an openly available ๐˜„๐—ผ๐—ฟ๐—น๐—ฑ๐˜„๐—ถ๐—ฑ๐—ฒ ๐—ฑ๐—ฎ๐˜๐—ฎ๐˜€๐—ฒ๐˜ ๐—ผ๐—ป ๐—ฟ๐—ผ๐—ฎ๐—ฑ ๐˜€๐˜‚๐—ฟ๐—ณ๐—ฎ๐—ฐ๐—ฒ ๐˜๐˜†๐—ฝ๐—ฒ๐˜€.

A new paper presents the methods behind the dataset.

โ–ซ๏ธ Studyยดs highlights: https://heigit.org/new-paper-paved-or-unpaved-a-deep-learning-derived-road-surface-global-dataset-from-mapillary-street-view-imagery-2/
โ–ซ๏ธ Full paper: https://www.sciencedirect.com/science/article/pii/S0924271625000784
โ–ซ๏ธ Dataset: https://data.humdata.org/organization/heidelberg-institute-for-geoinformation-technology

New Paper: โ€œPaved or unpaved? A Deep Learning derived Road Surface Global Dataset from Mapillary Street-View Imageryโ€ โ€“ HeiGIT

@heigit That's a very nice resource - any chance this could get added as a dataset to Rapid? :)
@Lumikeiju we are still investigating the best way to feed the data back to OSM in a process which is compatible with the community. We will come up with proposals to discuss within the OSM community soon