Skip to content

Latest commit

 

History

History
36 lines (27 loc) · 1.11 KB

DATA_PREPAER.md

File metadata and controls

36 lines (27 loc) · 1.11 KB

Data Preparation

Download

First download the text data and image data from the following links:

Data Link
Text Hugging Face HuggingFace
Image Google Drive Google Drive

Then create the data folder and organize the data as follows:

data/
├── nuscenes/
│   └── samples/
├── corruption/
│   ├── BitError/
│   ├── Bright/
│   ├── CameraCrash/
│   └── ...
├── drivebench-test.json
├── biterror.json
├── bright.json
└── camcrash.json

Preprocess

Preprocess the data by running the following command:

python tools/preprocess.py data/drivebench-test.json data/drivebench-test-final.json