Description: The considerable variety of indoor and outdoor waste materials scenarios was meticulously captured in our data collection and annotation methods. The majority of the thrash scene photographs were taken in West Bengal, India, in both urban and natural settings such as streets, parks, and lawns. We purposefully carried out this work in inclement weather, such as rain, and sunshine, as well as various daylight conditions, like morning, evening, and night. 6000 high-resolution photos with manually annotated ground truth for waste segmentation are included in the collection. The data was annotated using the LabelImg[1] program in both Pascal VOC and YOLO formats. The Dataset contains 21 classes such as Plastic carry, paper, paper_cup, plastic, plastic_coantainer, paper_cardboard, cigarette_packet, plastic_bottle, newspaper, thermocol, match_Box, metal, cloth, plastic_sack, medicine_wrapper, glass_bottle, glass, nylon, cotton, aluminium_foil and rubber with total 15884 Instances. Our dataset is available in
VOC Format: https://drive.google.com/file/d/1CeRmuX3jF3FKb6yKKN4ajgbnGpNQ8FYW/view?usp=share_link
YOLO Format: https://drive.google.com/file/d/1-0VIKv3rxQqaiIGFOQDrUsijoHD6Oe16/view?usp=share_link
If you use IURWDD in a scientific publication, we would appreciate using the following citations:
@deta{ak_iurwdd_2023,
year = {2023},
author = {Ali, Asfak and Sk, Md. Manarul and Pal, Debanka, aifuddin and Sinha Chaudhuri, Sheli}
title = {IURWDD: Indian Urban and Rural Waste Detection Dataset and Benchmark},
url = {https://github.com/asfakali/IURWDD.git}
}