Re-thinking face detection in the Covid-19 age | OpenCV

Face masks are now mandatory in the whole territory of Lombardia Area in the north of Italy and in many other areas worldwide. In an effort to contain the Covid-19 Coronavirus spread that has caused thousands of deaths, the local governments are insisting that millions of residents must wear protective face covering when they go out in public.

A funny consequence to covering their faces it’s the face masks trip up facial recognition functions, the technology necessary for many routine transactions in many country worldwide. Despite how much the face recognition is diffused a lot of experience such as certain mobile phones, entertainment apps and bank accounts won’t unlock anymore.

In January 2020, a Chinese company says it has developed the country’s first facial recognition technology that can identify people wearing a mask, as most are these days because of the #COVID19, and help in the fight against the disease.

The recognition rate can reach about 95%, which can ensure a quite good performance, the success rate for people without mask is estimated in 99.5%.

This Beijing-based firm used core technology developed over the past 10 years, a sample database of about a few million unmasked faces and a much smaller database of masked faces. It’s easy to argue how the new dataset has been inferred using a pretty easy to understand data augmentation technique: face with facial points, pick a random mask up, let’s project the mask according to the facial points = a (billion of) brand new masked faces.

In the same days someone teaches Iphone how to recognize his face wearing a mask. More: https://www.independent.co.uk/life-style/gadgets-and-tech/iphone-face-mask-id-recognition-unlock-coronavirus-covid-19-hack-a9459901.html

Most of the standard model in literature exploit information basically from eyes, mouth and face shape. And now? We starting figuring out how the face detection step should be modified / improved in order to handle with the missing information (occluded mouth and nose). So we developed the first LBP cascade, suitable for OpenCV, able to detect faces with masks. The cascade is totally free for academic, pro-bono or healthcare projects.

Masked faces LBP HAAR HOG OpenCV cascade

Full frontal (with partial profiles) masked face detection cascade, trained with:

  • ~7,000 positive samples (randomly sampled)
  • approx 0.9B of negative sub-regions containing outdoor and indoor samples (30%-70%)
  • Training size w=40 h=50 (aspect ratio 0.8)

LBP: it’s free for academic, pro-bono or healthcare projects. Send us an email here to get the link for the cascade!

  • Features set: 106080 features
  • Training time: ~2 days
  • TP: ~ 95.01% of positive training set
  • FN: ~ 04.99% of positive training set
  • FP: ~ 5.4e-006% of negative training set

HOG: (upon request)

HAAR: (upon request)

32 comments

  1. Hello,

    I am Daniel from CETUC/PUC-RIO. I have already published in linkedin about ArgoVision and Vision-Ary, received the NDA and sent it back signed, but have not received the LBP cascades yet.
    I think my response may have been filtered by some spam filter or lost in a lot of messages.
    If necessary I can send again the signed document. Thank you.

  2. Hello, I am a final year student in the Department of Computer Science and Engineering at Khulna University of Engineering &Technology, Bangladesh.
    For my academic thesis, I am doing research on the people traveling to the office by wearing masks during COVID-19. I am trying to develop a similar model. For a benchmark comparison of my developed model, I needed a reference model. The model, MASKED FACES LBP HAAR HOG OPENCV CASCADE is a perfect model for benchmark comparison.
    I had sent a response message earlier about the model request. But, I think my message has been filtered thus commenting on the website. Kindly please help me out.
    Thanks.

    Regards
    Shadmaan Hye

  3. Hello, for my academic thesis, I am doing research on the people traveling to the office by wearing masks during COVID-19. I am trying to develop a similar model. For a benchmark comparison of my developed model, I needed a reference model. The model, MASKED FACES LBP HAAR HOG OPENCV CASCADE is a perfect model for benchmark comparison.
    I had sent a response message earlier about the model request. But, I think my message has been filtered thus commenting on the website. Kindly please help me out.
    Thanks.

    Regards
    Shadmaan Hye

  4. Dear Team Vision ary,
    we want to try this haar for our research for covid screening using webcam and thermal camera, again we greatly appreciate your work.
    Thank you.

  5. Hi!
    I am a student of an IT course in Italy.
    For the final test I’m making an Android App wich recognise face from android’s camera smartphone.
    I have already achieve my goal to recognise faces and eyes, but now, with COVID-19 it will be great to recognise persons with mask.
    So I found your new Cascade and I will be very thankfully if you can send it to me.
    Many thanks in advance.
    Eric.

  6. Hello,
    I need the haar_cascade for my masters graduation project in art school. Can you send it to me too? Thank you so much!
    Florian D.

  7. I tried it a few time too already, but i am not sure if my the contact form was working. Can i also get your sample to use it in my master graduations project?

    best wishes,
    Florian

  8. Hello I’m student from Asia Pacific University. I would like to try out the cascade in development of FYP. Can u email me the cascade?

  9. Hello im a university student currently developing face mask detection system can u pls send me the cascade to tryout, ty very much.

  10. Hi, I am doing a research project on Facial Recognition and this would help massively! Could you please email me the link to download this Cascade, Thank you.

Lascia un commento

Il tuo indirizzo email non sarΓ  pubblicato.