AzureTipsAndTricks/tip369.md в main

6.5 KiB

Исходник Постоянная ссылка Ответственный История

type

title

excerpt

How to identify faces with the Azure Face service

Automatic face detection

Detecting and recognizing faces programmatically can be very difficult. Azure Face service provides this capability to you as-a-service. Azure Face service is an API that you can call and use in your applications to detect faces and facial features in images. It comes with a pre-trained machine learning model, so you only have to use the service, not configure or manage it.

In this post, we'll use Azure Face service to detect a face and facial features.

Prerequisites

If you want to follow along, you'll need the following:

An Azure subscription (If you don't have an Azure subscription, create a free account before you begin)

Create and use an Azure Cognitive Services Face Service

We'll start by creating a Face service resource in the Azure portal.

Go to the Azure portal
Click the Create a resource button (the plus-sign in the top left corner)
Search for Face, select the "Face" result and click Create
1. Select a Resource Group
2. Pick a Region for the service
3. Fill in a Name
4. Select a Pricing tier. The "Free F0" tier is fine
5. Check the box to agree to the terms
6. Click Review + create and then Create

(Create a Face service resource)

When the Face service is created, navigate to it in the Azure portal.

In the Face service, click on the Quick start menu
We will use the API console to test the service. Click on API console
Scroll down and select the Azure region in which you created the service. This will reload the page and makes sure that we are calling the correct service endpoint
We will enter all the information to make an API call to the Face service and test it. Scroll to the Query parameters

(Query parameters in the API Console)

For returnFaceAttributes, enter "age,gender,headPose,smile,facialHair,glasses,emotion,hair,makeup,occlusion,accessories,blur,exposure,noise". You can find a list of all possible face attributes here
Change the detectionModel to detection_01
Go to Headers. You need to enter the Face service subscription key here
1. Go back to the Face service in the Azure portal
2. Navigate to the Keys and Endpoint menu
3. Copy KEY 1

(Keys and endpoints of the Face service)

Back in the API console, paste the key value in the Ocp-Apim-Subscription-Key field
In the request body, change the url value to "https://upload.wikimedia.org/wikipedia/commons/c/c3/RH_Louise_Lillian_Gish.jpg". This is a sample picture on wikipedia

(Sample picture on wikipedia)

Click Send to send the request to the Face service
The Response status should be 200 OK if everything went alright. This should result in a Response content that looks like the one below. This shows that one face was detected in the image, and that the face is likely to be of a female, 23 years old, without glasses and with brown hair. Pretty cool, right?

x-envoy-upstream-service-time: 533
apim-request-id: c35cf15b-d1ac-4b36-9c13-aaf758daaff4
Strict-Transport-Security: max-age=31536000; includeSubDomains; preload
x-content-type-options: nosniff
CSP-Billing-Usage: CognitiveServices.Face.Transaction=1
Date: Fri, 07 15:13:42 GMT
Content-Length: 1019
Content-Type: application/json; charset=utf-8

[{
  "faceId": "be10f68d-b895-40b6-b3bb-e57bde7fcd9c",
  "faceRectangle": {
    "top": 131,
    "left": 177,
    "width": 162,
    "height": 162
  },
  "faceAttributes": {
    "smile": 0.001,
    "headPose": {
      "pitch": -5.7,
      "roll": -9.1,
      "yaw": -34.1
    },
    "gender": "female",
    "age": 23.0,
    "facialHair": {
      "moustache": 0.0,
      "beard": 0.0,
      "sideburns": 0.0
    },
    "glasses": "NoGlasses",
    "emotion": {
      "anger": 0.0,
      "contempt": 0.0,
      "disgust": 0.0,
      "fear": 0.0,
      "happiness": 0.001,
      "neutral": 0.987,
      "sadness": 0.001,
      "surprise": 0.01
    },
    "blur": {
      "blurLevel": "low",
      "value": 0.06
    },
    "exposure": {
      "exposureLevel": "goodExposure",
      "value": 0.67
    },
    "noise": {
      "noiseLevel": "low",
      "value": 0.0
    },
    "makeup": {
      "eyeMakeup": true,
      "lipMakeup": true
    },
    "accessories": [],
    "occlusion": {
      "foreheadOccluded": false,
      "eyeOccluded": false,
      "mouthOccluded": false
    },
    "hair": {
      "bald": 0.01,
      "invisible": false,
      "hairColor": [{
        "color": "brown",
        "confidence": 1.0
      }, {
        "color": "gray",
        "confidence": 0.53
      }, {
        "color": "black",
        "confidence": 0.51
      }, {
        "color": "blond",
        "confidence": 0.13
      }, {
        "color": "red",
        "confidence": 0.13
      }, {
        "color": "other",
        "confidence": 0.02
      }, {
        "color": "white",
        "confidence": 0.0
      }]
    }
  }
}]

Conclusion

You can use the Azure Face service to detect faces and facial features, just by calling an API. Go and check it out!

6.5 KiB Исходник Постоянная ссылка Ответственный История