Mapillary
Step 1. Download Mapillary dataset
Step 2. Unpack archive
Step 3. Directory structure have to be the following:
.├── training│ ├── images│ │ ├── 0035fkbjWljhaftpVM37-g.jpg│ │ ├── 00qclUcInksIYnm19b1Xfw.jpg│ │ ├── ...│ ├── instances│ │ ├── 0035fkbjWljhaftpVM37-g.png│ │ ├── 00qclUcInksIYnm19b1Xfw.png│ │ ├── ...│ └── labels│ ├── 0035fkbjWljhaftpVM37-g.png│ ├── 00qclUcInksIYnm19b1Xfw.png│ ├── ...└── validation├── images│ ├── 03G8WHFnNfiJR-457i0MWQ.jpg│ ├── 03T1YjbnUSumHU5n-iYEzA.jpg│ ├── ...├── instances│ ├── 03G8WHFnNfiJR-457i0MWQ.png│ ├── 03T1YjbnUSumHU5n-iYEzA.png│ ├── ...└── labels├── 03G8WHFnNfiJR-457i0MWQ.png├── 03T1YjbnUSumHU5n-iYEzA.png├── ...
Step 4. Choose all (two) subdirectories (training and validation) and drag and drop them to browser.
If you will drag and drop parent directory instead of its contents, import will crash.
Last updated 1 year ago