Computer Vision: The Next Era of Smart Policy

Computer Vision: The Next Era of Smart Policy

Hi there, my name is Ravit Sharma, and I’m CTO at Konect. At Konect (https://www.konect-co.com/), we've developed video footage analysis technology SpaSpect to monitor social distancing and face-covering regulation.

Take a second to put yourself in the shoes of a public health policymaker at a municipality. Your job is to make policy that curbs the spread of COVID-19 within your city. For example, you may decide to close the city’s shops and bars if there is a high risk of spread. The question is, how would you know when is the right time to make strategic policy decisions without unnecessarily hurting the economy?

Well, your job is hard for two reasons. First, you would be forced to consult indicators on the current status of the spread of the disease, which is currently limited to new cases of COVID-19 and ICU occupation. Because symptoms take 2-14 days to show up and not everyone is tested, these two indicators are both delayed and indirect measures. Secondly, even if you do take the appropriate action, you simply can’t reliably and confidently pinpoint how effective the regulation is without investing significant human resources across the city to observe whether social distancing and face-covering regulations are followed.

We’ve designed SpaSpect with a goal to address these current limitations. Our software analyzes video footage from security cameras overseeing public venues to continuously and accurately gather information such as

  • The proportion of people who are wearing masks and social distancing
  • How long visitors stay in the venue
  • Average distance between people
  • Factors that contribute to the greatest risk for spread

We also offer optional features, such as cough counting, contingent on the presence of a microphone

By periodically reporting this data to you, the public health manager, you can be sure that your decisions and regulations are rooted in pertinent and reliable data. Further, you can immediately assess the effectiveness of new policies and determine whether better enforcement is needed in a particular part of the city, such as the city parks.

Here’s a sample of what our realtime dashboard looks like where you can have a look at statistics of different areas in real time, in addition to periodic reports that are sent out.

Already, we're working with Bosch integration tools and are coordinating with counties and cities across the nation including the City of Southlake, Texas. Our goal is to maximize the potential of our software to businesses and local governments, and we're thrilled to keep working towards that goal!

要查看或添加评论,请登录

Ravit Sharma的更多文章

  • How I Improved the Performance of my Computer Vision Model Two-Fold

    How I Improved the Performance of my Computer Vision Model Two-Fold

    Python is great for training deep learning models. The variety of supported platforms make it easy for pretty much…

  • TensorFlow vs PyTorch

    TensorFlow vs PyTorch

    PyTorch and Tensorflow are two deep learning frameworks used in the creation of neural networks; they have largely the…

  • Object Tracking Methods

    Object Tracking Methods

    Object tracking is what it sounds like: the process of keeping track of an object as it moves across the screen. This…

  • Intel Vision Technology and OpenVINO

    Intel Vision Technology and OpenVINO

    Recently, I had the opportunity of attending a workshop of Intel's 2019 Embedded Vision Summit. The event was part of a…

  • Synthetic Gradients in Neural Networks

    Synthetic Gradients in Neural Networks

    A neural network is simply a complex function that processes input and spits out an output. Through computation of a…

  • UPGMA Method: Designing a Phylogenetic Tree

    UPGMA Method: Designing a Phylogenetic Tree

    A phylogenetic tree (AKA cladogram) is a diagrammatic representation of the evolutionary relatedness between various…

社区洞察

其他会员也浏览了