Live Face Recognition

ptangn 21 Reputation points

Dear team,

As part of the project requirement, I develop the system to perform "Live face recognition".

I already create the face API

I am quite new to this area, could we please advise me what to do to perform live face recognition?

Thank you so much in advance.

Best Regards

Peter Nguyen

0 comments No comments
Sign in to comment

Answer accepted by question author

SRILAKSHMI C 19,110 Reputation points • Microsoft External Staff • Moderator

Hello ptangn,

Welcome to Microsoft Q&A and Thank you for reaching out.

I understand that you're diving into "Live Face Recognition," and you're off to a great start by creating the Face API. Performing live face recognition typically involves capturing video, detecting faces in the frames, and recognizing those faces in real-time.

Here's what you can do to get started with live face recognition using Azure's services:

Set Up Live Video Analytics: You can leverage Azure's Live Video Analytics on IoT Edge. This allows you to plug in video analysis modules and analyze live video streams. Check out the Quickstart: Analyze live video with your own model for a detailed guide on how to set this up.

Integrate Face API:

Use the Azure Face API to detect faces in the frames captured from the live video feed. The Face API will return bounding box coordinates and recognize faces using its algorithms.

Make sure you're familiar with the parameters of the API, like recognitionModel and the types of face attributes you can retrieve. You can find more in the Face API documentation.

Process Frames:

Capture frames from the live video stream and send those to the Face API for detection and recognition. You might want to implement a filtering mechanism to process only some frames (e.g., every second frame) to optimize performance.

If you're using a language like C# or JavaScript, there are SDKs available to help you with calling the API easily. Check out the C# Quickstart and for examples.

Liveness Detection: If you need to ensure that the recognition is happening on a live person and to prevent spoofing, you could look into liveness detection. You'd have to fill out a form for access to the liveness feature through the Face Recognition intake form.

Testing and Privacy: Finally, do ensure that your implementation is compliant with privacy laws and always notify individuals when you're using their biometric data for recognition.

I hope this helps you get on the right track! Let me know if you have any questions or need help with specific aspects of your project.

Also, please refer this

I Hope this helps. Do let me know if you have any further queries.

If this answers your query, please do click Accept Answer and Yes for was this answer helpful.

Thank you!

  1. SRILAKSHMI C 19,110 Reputation points • Microsoft External Staff • Moderator

    Hi ptangn,

    Following up to see if the above answer was helpful. If this answers your query, do click Accept Answer and Yes for was this answer helpful. And, if you have any further query do let us know.

    Thank you!

  2. SRILAKSHMI C 19,110 Reputation points • Microsoft External Staff • Moderator

    Hi ptangn,

    Just checking in to see if you have got a chance to see my response to your question in resolving the issue.

    If you are still facing any further issues, please don't hesitate to reach out to us. We are happy to assist you.

    Looking forward to your response and appreciate your time on this.

    If you feel that your quires have been resolved, please accept the answer by clicking the "Upvote" and "Accept Answer" on the post.

    Thank you!

  3. ptangn 21 Reputation points

    Hi SRILASHMI C,

    Thank you for providing valuable instructions.

    I am currently waiting for "Azure AI Face API Recognition Limited Access Review" approval. Once, I have the access to the API, I will then dive into deeper.

    Thank you again.

Sign in to comment

2 additional answers

  1. ptangn 21 Reputation points

    We would like to implement the mechanism where the system will detect the existing images from live image

    0 comments No comments
    Sign in to comment

  2. AI answer

    To perform live face recognition using the Azure Face API, you will need to integrate facial liveness detection with face verification. Here are the steps you can follow:

    1. Select a Reference Image: Choose a reference image that will be used for verification. Ensure that the image meets the composition requirements for ID verification scenarios.
    2. Set Up Liveness Detection with Verification: You will need to orchestrate the liveness detection with the verification process. This involves:
      • Providing the verification reference image when creating a liveness session.
      • Using the appropriate SDK or REST API to create a session and initiate the liveness detection.
    3. Check Results: After the liveness session is created, you can query for the verification result along with the liveness result. This will confirm whether the person is physically present and matches the reference image.
    4. Delete Session: If you no longer need the session results, you can delete the session to clean up resources.

    For detailed guidance, you can refer to the Azure documentation on performing liveness detection with face verification. This documentation provides code examples and further instructions on how to implement these features across various platforms and languages.

    References:

    AI-generated content may be incorrect. Read our transparency notes for more information.
Sign in to answer

Your answer