IDD-AW: A Benchmark for Safe and Robust Segmentation of Drive Scenes in Unstructured Traffic and Adverse Weather
Complete Toolkit code, Dataset and Results will be available soon.
Paper link: https://shorturl.at/vwBG4
Website Link: https://iddaw.github.io/
The workflow for dataset download is:
- Open https://iddaw.github.io/
- Click on Dataset button
- (redirects to IDD site) https://idd.insaan.iiit.ac.in/dataset/download/
- Need to register and create a user
- Scroll down and click on IDD-AW link to get a 24 hour token to download the dataset.
- Click on the link and you can download the dataset.
For mask generation, we follow the same instructions as specified in the below repository for IDD Dataset : https://github.com/AutoNUE/public-code
However, just replace createLabels_iddaw.py in place of createLabels.py and run the same commands.