Clean/Littered Road Classification

The dataset contains images of clean and dirty road.

There are a total of 237 images, all of which bootstraped from the internet. The task is to create a classification model, which can accurately classify if a road is clean or littered. Because of the lack of data, pretrained models and data augmentation may be used.

Such a classification model can be used to develope applications to detect littered part of roads using cameras and send necessary service to those areas.

Naviagate Dataset:

1. Images: Folder containing all the road images.
2. metadata.csv: A csv file mapping the image name with the class label.