Simplify face recognition steps

Max.Bom 11 Reputation points
2020-10-27T02:47:42.943+00:00

Hello Azure,

Is it possible to make face recognition only need one time network I/O?

As I know if I want to do face recognition I need to do

  1. Send pic to do face detection. => Network I/O
  2. Send step one result to do face recognition => network I/O

Sending network request twice, is it possible to make it become once?

Because I want to reduce the whole process time.

Thanks.

Max

Azure Face
Azure Face
An Azure service that provides artificial intelligence algorithms that detect, recognize, and analyze human faces in images.
154 questions
Azure AI services
Azure AI services
A group of Azure services, SDKs, and APIs designed to make apps more intelligent, engaging, and discoverable.
2,415 questions
0 comments No comments
{count} vote

2 answers

Sort by: Most helpful
  1. Max.Bom 11 Reputation points
    2020-11-05T01:12:35.667+00:00

    @Ramr-msft Got it, thanks for replying.

    1 person found this answer helpful.

  2. Ramr-msft 17,616 Reputation points
    2020-10-28T07:22:45.097+00:00

    @98614511 Thanks for the question. Broadly, recognition describes the work of comparing two different faces to determine if they're similar or belong to the same person. Currently two API calls are required, you are using it most efficient way possible.

    1. See here the spec for Find Similar API. You can pass in a single face id, or an array of face ids, for the look up against a given large face list.
    2. See here the spec for Large Face List – Add Face. It only takes a single face at a time. It works in a similar way for large person groups as well.

    EDIT:

    @98614511 At this time we don't have an ability to combine detection and identification calls into one. This has been a common ask and we'll try to address with next major version of our service. For now, you are using it most efficient way possible.

    You can also raise a user voice request here so the community can vote and provide their feedback, the product team then checks this feedback and implements the feature in future releases.