Six steps of hand washing dataset with segmentation maps

Citation Author(s):
PRANAV
PRABHA
AMMACHI Labs
SREEJITH
SASIDHARAN
AMMACHI Labs
PRAVEEN
PANKAJAKSHAN
CropIn Labs
GAYATHRI
MANIKUTTY
AMMACHI Labs
Submitted by:
PRANAV PRABHA
Last updated:
Thu, 10/13/2022 - 21:47
DOI:
10.21227/ghr9-y726
License:
147 Views
Categories:
0
0 ratings - Please login to submit your rating.

Abstract 

 

The dataset contains raw frames of 7 classes, each denoting the different steps of handwashing. The dataset is created using the Hand Wash Dataset from Kaggle ( https://www.kaggle.com/datasets/realtimear/hand-wash-dataset ). Individual image frames were extracted from the original video dataset and cropped to 224*224 pixels. Bad images in the dataset, without handwashing steps or severe motion blurring, were removed from the dataset. Finally, we have created separate segmentation maps to mask the foreground images (hands) from the background images. The dataset will help conduct computer vision-based hand washing and hand hygiene research.

Instructions: 

The dataset can be used for research after giving the creators the required credit. We would like to acknowledge that we did not create the original video dataset used for creating this dataset. The link to the original dataset is https://www.kaggle.com/datasets/realtimear/hand-wash-dataset.

Comments

.

Submitted by Sohaib Asif on Sun, 04/23/2023 - 13:00