The authors of the Indian Roads dataset have created it for delivery robots with knee-height delivery robots. Sidewalk delivery robots can also use this dataset as it consists of side road data as well. The dataset consists of a semantic segmentation dataset of Indian roads for roads, potholes, footpaths, shallow paths, and backgrounds. The authors of the dataset have tested it using MobileNetv3 and ESPNetv2.

Semantic Segmentation Dataset of Indian Roads is a dataset for instance segmentation, semantic segmentation, and object detection tasks. Possible applications of the dataset could be in the automotive industry.

The dataset consists of 3227 images with 8129 labeled objects belonging to 4 different classes including road, footpath, shallow, and other: pothole.

Images in the Indian Roads dataset have pixel-level instance segmentation annotations. Due to the nature of the instance segmentation task, it can be automatically transformed into a semantic segmentation (only one mask for every class) or object detection (bounding boxes for every object) tasks. All images are labeled (i.e. with annotations). There are 2 splits in the dataset: train (2475 images) and val (752 images). The dataset was released in 2021 by the e-Yantra, IIT Bombay, India.

Semantic Segmentation Dataset of Indian Roads is under GNU GPL 2.0 license.

Source

If you make use of the Indian Roads data, please cite the following reference:

@dataset{Indian Roads,
  author={Simranjeet Singh and Soofiyan Atar and Aditya Panwar and Amit Kumar and Srijan Agrawal and Sravya G and Ravikumar Chaurasia and Shreyas Sule and Kavi Arya},
  title={Semantic Segmentation Dataset of Indian Roads},
  year={2021},
  url={https://www.kaggle.com/datasets/eyantraiit/semantic-segmentation-datasets-of-indian-roads}
}

Source

Dataset Indian Roads can be downloaded in Supervisely format:

As an alternative, it can be downloaded with dataset-tools package:

pip install --upgrade dataset-tools

… using following python code:

import dataset_tools as dtools

dtools.download(dataset='Indian Roads', dst_dir='~/dataset-ninja/')

Make sure not to overlook the python code example available on the Supervisely Developer Portal. It will give you a clear idea of how to effortlessly work with the downloaded dataset.

The data in original format can be downloaded here.