Measuring Social Distance in Images Using An AI- Human Team

EasyChair Preprint 4256

Abstract

Coronavirus was announced as a pandemic by the World Health Organization and affected almost all countries with millions of confirmed cases and hundreds of thousands of deaths. There is no vaccination or treatment to the date of writing this paper, and there are guidelines provided by the WHO and CDC to stay home and keep social distance while being out. Our work is focusing on social distance to help people staying safe and healthy and protect themselves from being infected by the virus. The system provides a search engine that offers information about the virus with image search that label images based on the social distance guideline. People can take precaution steps when they go out for shopping or working on essential role. We use information retrieval to fetch data from news agencies, augment these images from google images. We use AI service to determine if images have people, and human computation [1] via Amazon Mechanical Turk (MTurk) to help with labeling the images according to the social distance guide-lines. Users can check these images before they go to grocery stores or work.

Keyphrases: AI Human Team, Artificial Intelligence, COVID19, Coronavirus, MTurk, search engine, social distancing

@booklet{EasyChair:4256,
  author    = {Mohannad Al Ameedi},
  title     = {Measuring Social Distance in Images Using An AI- Human Team},
  howpublished = {EasyChair Preprint 4256},
  year      = {EasyChair, 2020}}