Plugins

Supervisely/ KITTI

Import Updated An hour ago Free
Import 'training' folder from KITTY Semantic Segmentation dataset
Free Signup

KITTI Import

The KITTI semantic segmentation benchmark consists of 200 semantically annotated train as well as 200 test images corresponding to the KITTI Stereo and Flow Benchmark 2015.

Usage steps:

1) Download KITTI dataset from official site.

2) Unpack archive

3) Directory structure have to be the following:

.
├── testing
│   └── image_2
│       ├── 000000_10.png
│       ├── ...
│       └── 000199_10.png
└── training
    ├── image_2
    │   ├── 000000_10.png
    │   ├── ...
    │   └── 000199_10.png
    ├── instance
    │   ├── 000000_10.png
    │   ├── ...
    │   └── 000199_10.png
    ├── semantic
    │   ├── 000000_10.png
    │   ├── ...
    │   └── 000199_10.png
    └── semantic_rgb
        ├── 000000_10.png
        ├── ...
        └── 000199_10.png

4) Open Supervisely import page. Choose KITTI import plugin.

5) Select training subdirectory and drag and drop them to browser. Wait a little bit.

6) Define new project name and click on START IMPORT button.

7) After import task finish, you can view project samples:

![](https://i.imgur.com/BsqqYcr.jpg)

![](https://i.imgur.com/PWXbZxJ.jpg)

8) On Statistics project tab you can get more information, for example - class distribution:

![](https://i.imgur.com/eQmXvBn.png)

Notes:

  • If you will drag and drop parent directory instead of its content, import will crash.
  • Supervisely import documentation.

Docker Image

supervisely/import-kitti:latest

More Info

ID
75269
Version ID
84487
First released
5 months ago
Last updated
An hour ago

Owner

s