# Label Images
<a name="sms-label-images"></a>

Use Ground Truth to label images. Select one of the following built in task types to learn more about that task type. Each page includes instructions to help you create a labeling job using that task type.

**Tip**  
To learn more about supported file types and input data quotas, see [Input data](sms-data-input.md).

**Topics**
+ [Classify image objects using a bounding box](sms-bounding-box.md)
+ [Identify image contents using semantic segmentation](sms-semantic-segmentation.md)
+ [Auto-Segmentation Tool](sms-auto-segmentation.md)
+ [Create an image classification job (Single Label)](sms-image-classification.md)
+ [Create an image classification job (Multi-label)](sms-image-classification-multilabel.md)
+ [Image Label Verification](sms-label-verification.md)

# Classify image objects using a bounding box
<a name="sms-bounding-box"></a>

The images used to train a machine learning model often contain more than one object. To classify and localize one or more objects within images, use the Amazon SageMaker Ground Truth bounding box labeling job task type. In this context, localization means the pixel-location of the bounding box. You create a bounding box labeling job using the Ground Truth section of the Amazon SageMaker AI console or the [https://docs.aws.amazon.com/sagemaker/latest/APIReference/API_CreateLabelingJob.html](https://docs.aws.amazon.com/sagemaker/latest/APIReference/API_CreateLabelingJob.html) operation.

**Important**  
For this task type, if you create your own manifest file, use `"source-ref"` to identify the location of each image file in Amazon S3 that you want labeled. For more information, see [Input data](sms-data-input.md).

## Creating a Bounding Box Labeling Job (Console)
<a name="sms-creating-bounding-box-labeling-job-console"></a>

You can follow the instructions [Create a Labeling Job (Console)](sms-create-labeling-job-console.md) to learn how to create a bounding box labeling job in the SageMaker AI console. In Step 10, choose **Image** from the **Task category** drop down menu, and choose **Bounding box** as the task type. 

Ground Truth provides a worker UI similar to the following for labeling tasks. When you create the labeling job with the console, you specify instructions to help workers complete the job and up to 50 labels that workers can choose from. 

![\[Gif showing how to draw a box around an object for a category.\]](http://docs.aws.amazon.com/sagemaker/latest/dg/images/sms/gifs/bb-sample.gif)


## Create a Bounding Box Labeling Job (API)
<a name="sms-creating-bounding-box-labeling-job-api"></a>

To create a bounding box labeling job, use the SageMaker API operation `CreateLabelingJob`. This API defines this operation for all AWS SDKs. To see a list of language-specific SDKs supported for this operation, review the **See Also** section of [https://docs.aws.amazon.com/sagemaker/latest/APIReference/API_CreateLabelingJob.html](https://docs.aws.amazon.com/sagemaker/latest/APIReference/API_CreateLabelingJob.html).

Follow the instructions on [Create a Labeling Job (API)](sms-create-labeling-job-api.md) and do the following while you configure your request: 
+ Pre-annotation Lambda functions for this task type end with `PRE-BoundingBox`. To find the pre-annotation Lambda ARN for your Region, see [PreHumanTaskLambdaArn](https://docs.aws.amazon.com/sagemaker/latest/dg/API_HumanTaskConfig.html#SageMaker-Type-HumanTaskConfig-PreHumanTaskLambdaArn) . 
+ Annotation-consolidation Lambda functions for this task type end with `ACS-BoundingBox`. To find the annotation-consolidation Lambda ARN for your Region, see [AnnotationConsolidationLambdaArn](https://docs.aws.amazon.com/sagemaker/latest/dg/API_AnnotationConsolidationConfig.html#SageMaker-Type-AnnotationConsolidationConfig-AnnotationConsolidationLambdaArn). 

The following is an example of an [AWS Python SDK (Boto3) request](https://boto3.amazonaws.com/v1/documentation/api/latest/reference/services/sagemaker.html#SageMaker.Client.create_labeling_job) to create a labeling job in the US East (N. Virginia) Region. All parameters in red should be replaced with your specifications and resources. 

```
response = client.create_labeling_job(
    LabelingJobName='example-bounding-box-labeling-job,
    LabelAttributeName='label',
    InputConfig={
        'DataSource': {
            'S3DataSource': {
                'ManifestS3Uri': 's3://bucket/path/manifest-with-input-data.json'
            }
        },
        'DataAttributes': {
            'ContentClassifiers': [
                'FreeOfPersonallyIdentifiableInformation'|'FreeOfAdultContent',
            ]
        }
    },
    OutputConfig={
        'S3OutputPath': 's3://bucket/path/file-to-store-output-data',
        'KmsKeyId': 'string'
    },
    RoleArn='arn:aws:iam::*:role/*,
    LabelCategoryConfigS3Uri='s3://bucket/path/label-categories.json',
    StoppingConditions={
        'MaxHumanLabeledObjectCount': 123,
        'MaxPercentageOfInputDatasetLabeled': 123
    },
    HumanTaskConfig={
        'WorkteamArn': 'arn:aws:sagemaker:region:*:workteam/private-crowd/*',
        'UiConfig': {
            'UiTemplateS3Uri': 's3://bucket/path/worker-task-template.html'
        },
        'PreHumanTaskLambdaArn': 'arn:aws:lambda:us-east-1:432418664414:function:PRE-BoundingBox',
        'TaskKeywords': [
            'Bounding Box',
        ],
        'TaskTitle': 'Bounding Box task',
        'TaskDescription': 'Draw bounding boxes around objects in an image',
        'NumberOfHumanWorkersPerDataObject': 123,
        'TaskTimeLimitInSeconds': 123,
        'TaskAvailabilityLifetimeInSeconds': 123,
        'MaxConcurrentTaskCount': 123,
        'AnnotationConsolidationConfig': {
            'AnnotationConsolidationLambdaArn': 'arn:aws:lambda:us-east-1:432418664414:function:ACS-BoundingBox'
          }
        },
    Tags=[
        {
            'Key': 'string',
            'Value': 'string'
        },
    ]
)
```

### Provide a Template for Bounding Box Labeling Jobs
<a name="sms-create-labeling-job-bounding-box-api-template"></a>

If you create a labeling job using the API, you must supply a worker task template in `UiTemplateS3Uri`. Copy and modify the following template. Only modify the [https://docs.aws.amazon.com/sagemaker/latest/dg/sms-creating-instruction-pages.html#sms-creating-quick-instructions](https://docs.aws.amazon.com/sagemaker/latest/dg/sms-creating-instruction-pages.html#sms-creating-quick-instructions), [https://docs.aws.amazon.com/sagemaker/latest/dg/sms-creating-instruction-pages.html#sms-creating-full-instructions](https://docs.aws.amazon.com/sagemaker/latest/dg/sms-creating-instruction-pages.html#sms-creating-full-instructions), and `header`. Upload this template to S3, and provide the S3 URI for this file in `UiTemplateS3Uri`.

```
<script src="https://assets.crowd.aws/crowd-html-elements.js"></script>
<crowd-form>
  <crowd-bounding-box
    name="boundingBox"
    src="{{ task.input.taskObject | grant_read_access }}"
    header="please draw box"
    labels="{{ task.input.labels | to_json | escape }}"
  >

    <full-instructions header="Bounding box instructions">
      <ol><li><strong>Inspect</strong> the image</li><li><strong>Determine</strong> 
      if the specified label is/are visible in the picture.</li>
      <li><strong>Outline</strong> each instance of the specified label in the image using the provided “Box” tool.</li></ol>
      <ul><li>Boxes should fit tight around each object</li>
      <li>Do not include parts of the object are overlapping or that cannot be seen, even though you think you can interpolate the whole shape.</li>
      <li>Avoid including shadows.</li>
      <li>If the target is off screen, draw the box up to the edge of the image.</li>    
    </full-instructions>
  
    <short-instructions>
      <h3><span style="color: rgb(0, 138, 0);">Good example</span></h3>
      <p>Enter description of a correct bounding box label and add images</p>
      <h3><span style="color: rgb(230, 0, 0);">Bad example</span></h3>
      <p>Enter description of an incorrect bounding box label and add images</p>
    </short-instructions>
  
  </crowd-bounding-box>
</crowd-form>
```

## Bounding Box Output Data
<a name="sms-bounding-box-output-data"></a>

Once you have created a bounding box labeling job, your output data will be located in the Amazon S3 bucket specified in the `S3OutputPath` parameter when using the API or in the **Output dataset location** field of the **Job overview** section of the console. 

For example, the output manifest file of a successfully completed single-class bounding box task will contain the following: 

```
[
  {
    "boundingBox": {
      "boundingBoxes": [
        {
          "height": 2832,
          "label": "bird",
          "left": 681,
          "top": 599,
          "width": 1364
        }
      ],
      "inputImageProperties": {
        "height": 3726,
        "width": 2662
      }
    }
  }
]
```

The `boundingBoxes` parameter identifies the location of the bounding box drawn around an object identified as a "bird" relative to the top-left corner of the image which is taken to be the (0,0) pixel-coordinate. In the previous example, **`left`** and **`top`** identify the location of the pixel in the top-left corner of the bounding box relative to the top-left corner of the image. The dimensions of the bounding box are identified with **`height`** and **`width`**. The `inputImageProperties` parameter gives the pixel-dimensions of the original input image.

When you use the bounding box task type, you can create single- and multi-class bounding box labeling jobs. The output manifest file of a successfully completed multi-class bounding box will contain the following: 

```
[
  {
    "boundingBox": {
      "boundingBoxes": [
        {
          "height": 938,
          "label": "squirrel",
          "left": 316,
          "top": 218,
          "width": 785
        },
        {
          "height": 825,
          "label": "rabbit",
          "left": 1930,
          "top": 2265,
          "width": 540
        },
        {
          "height": 1174,
          "label": "bird",
          "left": 748,
          "top": 2113,
          "width": 927
        },
        {
          "height": 893,
          "label": "bird",
          "left": 1333,
          "top": 847,
          "width": 736
        }
      ],
      "inputImageProperties": {
        "height": 3726,
        "width": 2662
      }
    }
  }
]
```

To learn more about the output manifest file that results from a bounding box labeling job, see [Bounding box job output](sms-data-output.md#sms-output-box).

To learn more about the output manifest file generated by Ground Truth and the file structure the Ground Truth uses to store your output data, see [Labeling job output data](sms-data-output.md). 

# Identify image contents using semantic segmentation
<a name="sms-semantic-segmentation"></a>

To identify the contents of an image at the pixel level, use an Amazon SageMaker Ground Truth semantic segmentation labeling task. When assigned a semantic segmentation labeling job, workers classify pixels in the image into a set of predefined labels or classes. Ground Truth supports single and multi-class semantic segmentation labeling jobs. You create a semantic segmentation labeling job using the Ground Truth section of the Amazon SageMaker AI console or the [https://docs.aws.amazon.com/sagemaker/latest/APIReference/API_CreateLabelingJob.html](https://docs.aws.amazon.com/sagemaker/latest/APIReference/API_CreateLabelingJob.html) operation. 

Images that contain large numbers of objects that need to be segmented require more time. To help workers (from a private or vendor workforce) label these objects in less time and with greater accuracy, Ground Truth provides an AI-assisted auto-segmentation tool. For information, see [Auto-Segmentation Tool](sms-auto-segmentation.md).

**Important**  
For this task type, if you create your own manifest file, use `"source-ref"` to identify the location of each image file in Amazon S3 that you want labeled. For more information, see [Input data](sms-data-input.md).

## Creating a Semantic Segmentation Labeling Job (Console)
<a name="sms-creating-ss-labeling-job-console"></a>

You can follow the instructions [Create a Labeling Job (Console)](sms-create-labeling-job-console.md) to learn how to create a semantic segmentation labeling job in the SageMaker AI console. In Step 10, choose **Image** from the **Task category** drop down menu, and choose **Semantic segmentation** as the task type. 

Ground Truth provides a worker UI similar to the following for labeling tasks. When you create the labeling job with the console, you specify instructions to help workers complete the job and labels that workers can choose from. 

![\[Gif showing an example on how to create a semantic segmentation labeling job in the SageMaker AI console.\]](http://docs.aws.amazon.com/sagemaker/latest/dg/images/semantic_segmentation_sample.gif)


## Create a Semantic Segmentation Labeling Job (API)
<a name="sms-creating-ss-labeling-job-api"></a>

To create a semantic segmentation labeling job, use the SageMaker API operation `CreateLabelingJob`. This API defines this operation for all AWS SDKs. To see a list of language-specific SDKs supported for this operation, review the **See Also** section of [https://docs.aws.amazon.com/sagemaker/latest/APIReference/API_CreateLabelingJob.html](https://docs.aws.amazon.com/sagemaker/latest/APIReference/API_CreateLabelingJob.html).

Follow the instructions on [Create a Labeling Job (API)](sms-create-labeling-job-api.md) and do the following while you configure your request: 
+ Pre-annotation Lambda functions for this task type end with `PRE-SemanticSegmentation`. To find the pre-annotation Lambda ARN for your Region, see [PreHumanTaskLambdaArn](https://docs.aws.amazon.com/sagemaker/latest/dg/API_HumanTaskConfig.html#SageMaker-Type-HumanTaskConfig-PreHumanTaskLambdaArn) . 
+ Annotation-consolidation Lambda functions for this task type end with `ACS-SemanticSegmentation`. To find the annotation-consolidation Lambda ARN for your Region, see [AnnotationConsolidationLambdaArn](https://docs.aws.amazon.com/sagemaker/latest/dg/API_AnnotationConsolidationConfig.html#SageMaker-Type-AnnotationConsolidationConfig-AnnotationConsolidationLambdaArn). 

The following is an example of an [AWS Python SDK (Boto3) request](https://boto3.amazonaws.com/v1/documentation/api/latest/reference/services/sagemaker.html#SageMaker.Client.create_labeling_job) to create a labeling job in the US East (N. Virginia) Region. All parameters in red should be replaced with your specifications and resources. 

```
response = client.create_labeling_job(
    LabelingJobName='example-semantic-segmentation-labeling-job,
    LabelAttributeName='label',
    InputConfig={
        'DataSource': {
            'S3DataSource': {
                'ManifestS3Uri': 's3://bucket/path/manifest-with-input-data.json'
            }
        },
        'DataAttributes': {
            'ContentClassifiers': [
                'FreeOfPersonallyIdentifiableInformation'|'FreeOfAdultContent',
            ]
        }
    },
    OutputConfig={
        'S3OutputPath': 's3://bucket/path/file-to-store-output-data',
        'KmsKeyId': 'string'
    },
    RoleArn='arn:aws:iam::*:role/*,
    LabelCategoryConfigS3Uri='s3://bucket/path/label-categories.json',
    StoppingConditions={
        'MaxHumanLabeledObjectCount': 123,
        'MaxPercentageOfInputDatasetLabeled': 123
    },
    HumanTaskConfig={
        'WorkteamArn': 'arn:aws:sagemaker:region:*:workteam/private-crowd/*',
        'UiConfig': {
            'UiTemplateS3Uri': 's3://bucket/path/worker-task-template.html'
        },
        'PreHumanTaskLambdaArn': 'arn:aws:lambda:us-east-1:432418664414:function:PRE-SemanticSegmentation,
        'TaskKeywords': [
            'Semantic Segmentation',
        ],
        'TaskTitle': 'Semantic segmentation task',
        'TaskDescription': 'For each category provided, segment out each relevant object using the color associated with that category',
        'NumberOfHumanWorkersPerDataObject': 123,
        'TaskTimeLimitInSeconds': 123,
        'TaskAvailabilityLifetimeInSeconds': 123,
        'MaxConcurrentTaskCount': 123,
        'AnnotationConsolidationConfig': {
            'AnnotationConsolidationLambdaArn': 'arn:aws:lambda:us-east-1:432418664414:function:ACS-SemanticSegmentation'
        },
    Tags=[
        {
            'Key': 'string',
            'Value': 'string'
        },
    ]
)
```

### Provide a Template for Semantic Segmentation Labeling Jobs
<a name="sms-create-labeling-job-ss-api-template"></a>

If you create a labeling job using the API, you must supply a worker task template in `UiTemplateS3Uri`. Copy and modify the following template. Only modify the [https://docs.aws.amazon.com/sagemaker/latest/dg/sms-creating-instruction-pages.html#sms-creating-quick-instructions](https://docs.aws.amazon.com/sagemaker/latest/dg/sms-creating-instruction-pages.html#sms-creating-quick-instructions), [https://docs.aws.amazon.com/sagemaker/latest/dg/sms-creating-instruction-pages.html#sms-creating-full-instructions](https://docs.aws.amazon.com/sagemaker/latest/dg/sms-creating-instruction-pages.html#sms-creating-full-instructions), and `header`. 

Upload this template to S3, and provide the S3 URI for this file in `UiTemplateS3Uri`.

```
<script src="https://assets.crowd.aws/crowd-html-elements.js"></script>
<crowd-form>
  <crowd-semantic-segmentation
    name="crowd-semantic-segmentation"
    src="{{ task.input.taskObject | grant_read_access }}"
    header="Please segment out all pedestrians."
    labels="{{ task.input.labels | to_json | escape }}"
  >
    <full-instructions header="Segmentation instructions">
      <ol><li><strong>Read</strong> the task carefully and inspect the image.</li>
      <li><strong>Read</strong> the options and review the examples provided to understand more about the labels.</li>
      <li><strong>Choose</strong> the appropriate label that best suits an object and paint that object using the tools provided.</li></ol>
    </full-instructions>
    <short-instructions>
      <h2><span style="color: rgb(0, 138, 0);">Good example</span></h2>
      <p>Enter description to explain a correctly done segmentation</p>
      <p><br></p><h2><span style="color: rgb(230, 0, 0);">Bad example</span></h2>
      <p>Enter description of an incorrectly done segmentation</p>
    </short-instructions>
  </crowd-semantic-segmentation>
</crowd-form>
```

## Semantic Segmentation Output Data
<a name="sms-ss-ouput-data"></a>

Once you have created a semantic segmentation labeling job, your output data will be located in the Amazon S3 bucket specified in the `S3OutputPath` parameter when using the API or in the **Output dataset location** field of the **Job overview** section of the console. 

To learn more about the output manifest file generated by Ground Truth and the file structure the Ground Truth uses to store your output data, see [Labeling job output data](sms-data-output.md). 

To see an example of an output manifest file for a semantic segmentation labeling job, see [3D point cloud semantic segmentation output](sms-data-output.md#sms-output-point-cloud-segmentation).

# Auto-Segmentation Tool
<a name="sms-auto-segmentation"></a>

Image segmentation is the process of dividing an image into multiple segments, or sets of labeled pixels. In Amazon SageMaker Ground Truth, the process of identifying all pixels that fall under a given label involves applying a colored filler, or "mask", over those pixels. Some labeling job tasks contain images with a large numbers of objects that need to be segmented. To help workers label these objects in less time and with greater accuracy, Ground Truth provides an auto-segmentation tool for segmentation tasks assigned to private and vendor workforces. This tool uses a machine learning model to automatically segment individual objects in the image with minimal worker input. Workers can refine the mask generated by the auto-segmentation tool using other tools found in the worker console. This helps workers complete image segmentation tasks faster and more accurately, resulting in lower cost and higher label quality. The following page gives information about the tool and its availability.

**Note**  
The auto-segmentation tool is available for segmentation tasks that are sent to a private workforce or vendor workforce. It isn't available for tasks sent to the public workforce (Amazon Mechanical Turk). 

## Tool Preview
<a name="sms-auto-segment-tool-preview"></a>

When workers are assigned a labeling job that provides the auto-segmentation tool, they are provided with detailed instructions on how to use the tool. For example, a worker might see the following in the worker console: 

![\[Example UI with instructions on how to use the tool in the worker console.\]](http://docs.aws.amazon.com/sagemaker/latest/dg/images/sms/gifs/semantic-segmentation.gif)


Workers can use **View full instructions** to learn how to use the tool. Workers will need to place a point on four extreme-points ( top-most, bottom-most, left-most, and right-most points ) of the object of interest, and the tool will automatically generate a mask for the object. Workers can further-refine the mask using the other tools provided, or by using the auto-segment tool on smaller portions of the object that were missed. 

## Tool Availability
<a name="sms-auto-segment-tool-availability"></a>

The auto-segmentation tool automatically appears in your workers' consoles if you create a semantic segmentation labeling job using the Amazon SageMaker AI console. While creating a semantic segmentation job in the SageMaker AI console, you will be able to preview the tool while creating worker instructions. To learn how to create a semantic segmentation labeling job in the SageMaker AI console, see [Getting started: Create a bounding box labeling job with Ground Truth](sms-getting-started.md). 

If you are creating a custom instance segmentation labeling job in the SageMaker AI console or creating an instance- or semantic-segmentation labeling job using the Ground Truth API, you need to create a custom task template to design your worker console and instructions. To include the auto-segmentation tool in your worker console, ensure that the following conditions are met in your custom task template:
+ For semantic segmentation labeling jobs created using the API, the `<crowd-semantic-segmentation>` is present in the task template. For custom instance segmentation labeling jobs, the `<crowd-instance-segmentation>` tag is present in the task template.
+ The task is assigned to a private workforce or vendor workforce. 
+ The images to be labeled are Amazon Simple Storage Service Amazon S3) objects that have been pre-signed for the Worker so that they can access it. This is true if the task template includes the `grant_read_access` filter. For information about the `grant_read_access` filter, see [Adding automation with Liquid](sms-custom-templates-step2-automate.md).

The following is an example of a custom task template for a custom instance segmentation labeling job, which includes the `<crowd-instance-segmentation/>` tag and the `grant_read_access` Liquid filter.

```
<script src="https://assets.crowd.aws/crowd-html-elements.js"></script>
<crowd-form>
  <crowd-instance-segmentation
    name="crowd-instance-segmentation"
    src="{{ task.input.taskObject | grant_read_access }}"
    labels="['Car','Road']"
   <full-instructions header="Segmentation instructions">
      Segment each instance of each class of objects in the image. 
    </full-instructions>

    <short-instructions>
      <p>Segment each instance of each class of objects in the image.</p>

      <h3 style="color: green">GOOD EXAMPLES</h3>
      <img src="path/to/image.jpg" style="width: 100%">
      <p>Good because A, B, C.</p>

      <h3 style="color: red">BAD EXAMPLES</h3>
      <img src="path/to/image.jpg" style="width: 100%">
      <p>Bad because X, Y, Z.</p>
    </short-instructions>
  </crowd-instance-segmentation>
</crowd-form>
```

# Create an image classification job (Single Label)
<a name="sms-image-classification"></a>

Use an Amazon SageMaker Ground Truth image classification labeling task when you need workers to classify images using predefined labels that you specify. Workers are shown images and are asked to choose one label for each image. You can create an image classification labeling job using the Ground Truth section of the Amazon SageMaker AI console or the [https://docs.aws.amazon.com/sagemaker/latest/APIReference/API_CreateLabelingJob.html](https://docs.aws.amazon.com/sagemaker/latest/APIReference/API_CreateLabelingJob.html) operation. 

**Important**  
For this task type, if you create your own manifest file, use `"source-ref"` to identify the location of each image file in Amazon S3 that you want labeled. For more information, see [Input data](sms-data-input.md).

## Create an Image Classification Labeling Job (Console)
<a name="sms-creating-image-classification-console"></a>

You can follow the instructions [Create a Labeling Job (Console)](sms-create-labeling-job-console.md) to learn how to create a image classification labeling job in the SageMaker AI console. In Step 10, choose **Image** from the **Task category** drop down menu, and choose **Image Classification (Single Label)** as the task type. 

Ground Truth provides a worker UI similar to the following for labeling tasks. When you create the labeling job with the console, you specify instructions to help workers complete the job and labels that workers can choose from. 

![\[Example worker UI for labeling tasks, provided by Ground Truth.\]](http://docs.aws.amazon.com/sagemaker/latest/dg/images/image-classification-example.png)


## Create an Image Classification Labeling Job (API)
<a name="sms-creating-image-classification-api"></a>

To create an image classification labeling job, use the SageMaker API operation `CreateLabelingJob`. This API defines this operation for all AWS SDKs. To see a list of language-specific SDKs supported for this operation, review the **See Also** section of [https://docs.aws.amazon.com/sagemaker/latest/APIReference/API_CreateLabelingJob.html](https://docs.aws.amazon.com/sagemaker/latest/APIReference/API_CreateLabelingJob.html).

Follow the instructions on [Create a Labeling Job (API)](sms-create-labeling-job-api.md) and do the following while you configure your request: 
+ Pre-annotation Lambda functions for this task type end with `PRE-ImageMultiClass`. To find the pre-annotation Lambda ARN for your Region, see [PreHumanTaskLambdaArn](https://docs.aws.amazon.com/sagemaker/latest/dg/API_HumanTaskConfig.html#SageMaker-Type-HumanTaskConfig-PreHumanTaskLambdaArn) . 
+ Annotation-consolidation Lambda functions for this task type end with `ACS-ImageMultiClass`. To find the annotation-consolidation Lambda ARN for your Region, see [AnnotationConsolidationLambdaArn](https://docs.aws.amazon.com/sagemaker/latest/dg/API_AnnotationConsolidationConfig.html#SageMaker-Type-AnnotationConsolidationConfig-AnnotationConsolidationLambdaArn). 

The following is an example of an [AWS Python SDK (Boto3) request](https://boto3.amazonaws.com/v1/documentation/api/latest/reference/services/sagemaker.html#SageMaker.Client.create_labeling_job) to create a labeling job in the US East (N. Virginia) Region. All parameters in red should be replaced with your specifications and resources. 

```
response = client.create_labeling_job(
    LabelingJobName='example-image-classification-labeling-job',
    LabelAttributeName='label',
    InputConfig={
        'DataSource': {
            'S3DataSource': {
                'ManifestS3Uri': 's3://bucket/path/manifest-with-input-data.json'
            }
        },
        'DataAttributes': {
            'ContentClassifiers': [
                'FreeOfPersonallyIdentifiableInformation'|'FreeOfAdultContent',
            ]
        }
    },
    OutputConfig={
        'S3OutputPath': 's3://bucket/path/file-to-store-output-data',
        'KmsKeyId': 'string'
    },
    RoleArn='arn:aws:iam::*:role/*,
    LabelCategoryConfigS3Uri='s3://bucket/path/label-categories.json',
    StoppingConditions={
        'MaxHumanLabeledObjectCount': 123,
        'MaxPercentageOfInputDatasetLabeled': 123
    },
    HumanTaskConfig={
        'WorkteamArn': 'arn:aws:sagemaker:region:*:workteam/private-crowd/*',
        'UiConfig': {
            'UiTemplateS3Uri': 's3://bucket/path/worker-task-template.html'
        },
        'PreHumanTaskLambdaArn': 'arn:aws:lambda:us-east-1:432418664414:function:PRE-ImageMultiClass,
        'TaskKeywords': [
            Image classification',
        ],
        'TaskTitle': Image classification task',
        'TaskDescription': 'Carefully inspect the image and classify it by selecting one label from the categories provided.',
        'NumberOfHumanWorkersPerDataObject': 123,
        'TaskTimeLimitInSeconds': 123,
        'TaskAvailabilityLifetimeInSeconds': 123,
        'MaxConcurrentTaskCount': 123,
        'AnnotationConsolidationConfig': {
            'AnnotationConsolidationLambdaArn': 'arn:aws:lambda:us-east-1:432418664414:function:ACS-ImageMultiClass'
        },
    Tags=[
        {
            'Key': 'string',
            'Value': 'string'
        },
    ]
)
```

### Provide a Template for Image Classification Labeling Jobs
<a name="worker-template-image-classification"></a>

If you create a labeling job using the API, you must supply a worker task template in `UiTemplateS3Uri`. Copy and modify the following template. Only modify the [https://docs.aws.amazon.com/sagemaker/latest/dg/sms-creating-instruction-pages.html#sms-creating-quick-instructions](https://docs.aws.amazon.com/sagemaker/latest/dg/sms-creating-instruction-pages.html#sms-creating-quick-instructions), [https://docs.aws.amazon.com/sagemaker/latest/dg/sms-creating-instruction-pages.html#sms-creating-full-instructions](https://docs.aws.amazon.com/sagemaker/latest/dg/sms-creating-instruction-pages.html#sms-creating-full-instructions), and `header`. 

Upload this template to S3, and provide the S3 URI for this file in `UiTemplateS3Uri`.

```
<script src="https://assets.crowd.aws/crowd-html-elements.js"></script>
<crowd-form>
  <crowd-image-classifier
    name="crowd-image-classifier"
    src="{{ task.input.taskObject | grant_read_access }}"
    header="please classify"
    categories="{{ task.input.labels | to_json | escape }}"
  >
    <full-instructions header="Image classification instructions">
      <ol><li><strong>Read</strong> the task carefully and inspect the image.</li>
      <li><strong>Read</strong> the options and review the examples provided to understand more about the labels.</li>
      <li><strong>Choose</strong> the appropriate label that best suits the image.</li></ol>
    </full-instructions>
    <short-instructions>
      <h3><span style="color: rgb(0, 138, 0);">Good example</span></h3>
      <p>Enter description to explain the correct label to the workers</p>
      <h3><span style="color: rgb(230, 0, 0);">Bad example</span></h3><p>Enter description of an incorrect label</p>
    </short-instructions>
  </crowd-image-classifier>
</crowd-form>
```

## Image Classification Output Data
<a name="sms-image-classification-output-data"></a>

Once you have created an image classification labeling job, your output data will be located in the Amazon S3 bucket specified in the `S3OutputPath` parameter when using the API or in the **Output dataset location** field of the **Job overview** section of the console. 

To learn more about the output manifest file generated by Ground Truth and the file structure the Ground Truth uses to store your output data, see [Labeling job output data](sms-data-output.md). 

To see an example of an output manifest file from an image classification labeling job, see [Classification job output](sms-data-output.md#sms-output-class).

# Create an image classification job (Multi-label)
<a name="sms-image-classification-multilabel"></a>

Use an Amazon SageMaker Ground Truth multi-label image classification labeling task when you need workers to classify multiple objects in an image. For example, the following image features a dog and a cat. You can use multi-label image classification to associate the labels "dog" and "cat" with this image. The following page gives information about creating an image classification job.

![\[Photo by Anusha Barwa on Unsplash\]](http://docs.aws.amazon.com/sagemaker/latest/dg/images/dog-cat-photo.jpg)


When working on a multi-label image classification task, workers should choose all applicable labels, but must choose at least one. When creating a job using this task type, you can provide up to 50 label-categories. 

When creating a labeling job in the console, Ground Truth doesn't provide a "none" category for when none of the labels applies to an image. To provide this option to workers, include a label similar to "none" or "other" when you create a multi-label image classification job. 

To restrict workers to choosing a single label for each image, use the [Create an image classification job (Single Label)](sms-image-classification.md) task type.

**Important**  
For this task type, if you create your own manifest file, use `"source-ref"` to identify the location of each image file in Amazon S3 that you want labeled. For more information, see [Input data](sms-data-input.md).

## Create a Multi-Label Image Classification Labeling Job (Console)
<a name="sms-creating-multilabel-image-classification-console"></a>

You can follow the instructions [Create a Labeling Job (Console)](sms-create-labeling-job-console.md) to learn how to create a multi-label image classification labeling job in the SageMaker AI console. In Step 10, choose **Image** from the **Task category** drop down menu, and choose **Image Classification (Multi-label)** as the task type. 

Ground Truth provides a worker UI similar to the following for labeling tasks. When you create a labeling job in the console, you specify instructions to help workers complete the job and labels that workers can choose from. 

![\[Example worker UI for labeling tasks, provided by Ground Truth.\]](http://docs.aws.amazon.com/sagemaker/latest/dg/images/image-classification-multilabel-example.png)


## Create a Multi-Label Image Classification Labeling Job (API)
<a name="sms-create-multi-select-image-classification-job-api"></a>

To create a multi-label image classification labeling job, use the SageMaker API operation `CreateLabelingJob`. This API defines this operation for all AWS SDKs. To see a list of language-specific SDKs supported for this operation, review the **See Also** section of [https://docs.aws.amazon.com/sagemaker/latest/APIReference/API_CreateLabelingJob.html](https://docs.aws.amazon.com/sagemaker/latest/APIReference/API_CreateLabelingJob.html).

Follow the instructions on [Create a Labeling Job (API)](sms-create-labeling-job-api.md) and do the following while you configure your request: 
+ Pre-annotation Lambda functions for this task type end with `PRE-ImageMultiClassMultiLabel`. To find the pre-annotation Lambda ARN for your Region, see [PreHumanTaskLambdaArn](https://docs.aws.amazon.com/sagemaker/latest/dg/API_HumanTaskConfig.html#SageMaker-Type-HumanTaskConfig-PreHumanTaskLambdaArn) . 
+ Annotation-consolidation Lambda functions for this task type end with `ACS-ImageMultiClassMultiLabel`. To find the annotation-consolidation Lambda ARN for your Region, see [AnnotationConsolidationLambdaArn](https://docs.aws.amazon.com/sagemaker/latest/dg/API_AnnotationConsolidationConfig.html#SageMaker-Type-AnnotationConsolidationConfig-AnnotationConsolidationLambdaArn). 

The following is an example of an [AWS Python SDK (Boto3) request](https://boto3.amazonaws.com/v1/documentation/api/latest/reference/services/sagemaker.html#SageMaker.Client.create_labeling_job) to create a labeling job in the US East (N. Virginia) Region. All parameters in red should be replaced with your specifications and resources. 

```
response = client.create_labeling_job(
    LabelingJobName='example-multi-label-image-classification-labeling-job,
    LabelAttributeName='label',
    InputConfig={
        'DataSource': {
            'S3DataSource': {
                'ManifestS3Uri': 's3://bucket/path/manifest-with-input-data.json'
            }
        },
        'DataAttributes': {
            'ContentClassifiers': [
                'FreeOfPersonallyIdentifiableInformation'|'FreeOfAdultContent',
            ]
        }
    },
    OutputConfig={
        'S3OutputPath': 's3://bucket/path/file-to-store-output-data',
        'KmsKeyId': 'string'
    },
    RoleArn='arn:aws:iam::*:role/*,
    LabelCategoryConfigS3Uri='s3://bucket/path/label-categories.json',
    StoppingConditions={
        'MaxHumanLabeledObjectCount': 123,
        'MaxPercentageOfInputDatasetLabeled': 123
    },
    HumanTaskConfig={
        'WorkteamArn': 'arn:aws:sagemaker:region:*:workteam/private-crowd/*',
        'UiConfig': {
            'UiTemplateS3Uri': 's3://bucket/path/worker-task-template.html'
        },
        'PreHumanTaskLambdaArn': 'arn:aws:lambda:us-east-1:432418664414:function:PRE-ImageMultiClassMultiLabel',
        'TaskKeywords': [
            'Image Classification',
        ],
        'TaskTitle': 'Multi-label image classification task',
        'TaskDescription': 'Select all labels that apply to the images shown',
        'NumberOfHumanWorkersPerDataObject': 123,
        'TaskTimeLimitInSeconds': 123,
        'TaskAvailabilityLifetimeInSeconds': 123,
        'MaxConcurrentTaskCount': 123,
        'AnnotationConsolidationConfig': {
            'AnnotationConsolidationLambdaArn': 'arn:aws:lambda:us-east-1:432418664414:function:ACS-ImageMultiClassMultiLabel'
        },
    Tags=[
        {
            'Key': 'string',
            'Value': 'string'
        },
    ]
)
```

### Provide a Template for Multi-label Image Classification
<a name="sms-custom-template-multi-image-label-classification"></a>

If you create a labeling job using the API, you must supply a worker task template in `UiTemplateS3Uri`. Copy and modify the following template. Only modify the [https://docs.aws.amazon.com/sagemaker/latest/dg/sms-creating-instruction-pages.html#sms-creating-quick-instructions](https://docs.aws.amazon.com/sagemaker/latest/dg/sms-creating-instruction-pages.html#sms-creating-quick-instructions), [https://docs.aws.amazon.com/sagemaker/latest/dg/sms-creating-instruction-pages.html#sms-creating-full-instructions](https://docs.aws.amazon.com/sagemaker/latest/dg/sms-creating-instruction-pages.html#sms-creating-full-instructions), and `header`. 

Upload this template to S3, and provide the S3 URI for this file in `UiTemplateS3Uri`.

```
<script src="https://assets.crowd.aws/crowd-html-elements.js"></script>
<crowd-form>
  <crowd-image-classifier-multi-select
    name="crowd-image-classifier-multi-select"
    src="{{ task.input.taskObject | grant_read_access }}"
    header="Please identify all classes in image"
    categories="{{ task.input.labels | to_json | escape }}"
  >
    <full-instructions header="Multi Label Image classification instructions">
      <ol><li><strong>Read</strong> the task carefully and inspect the image.</li>
      <li><strong>Read</strong> the options and review the examples provided to understand more about the labels.</li>
       <li><strong>Choose</strong> the appropriate labels that best suit the image.</li></ol>
    </full-instructions>
    <short-instructions>
      <h3><span style="color: rgb(0, 138, 0);">Good example</span></h3>
      <p>Enter description to explain the correct label to the workers</p>
      <h3><span style="color: rgb(230, 0, 0);">Bad example</span></h3>
      <p>Enter description of an incorrect label</p>
   </short-instructions>
  </crowd-image-classifier-multi-select>
</crowd-form>
```

## Multi-label Image Classification Output Data
<a name="sms-image-classification-multi-output-data"></a>

Once you have created a multi-label image classification labeling job, your output data will be located in the Amazon S3 bucket specified in the `S3OutputPath` parameter when using the API or in the **Output dataset location** field of the **Job overview** section of the console. 

To learn more about the output manifest file generated by Ground Truth and the file structure the Ground Truth uses to store your output data, see [Labeling job output data](sms-data-output.md). 

To see an example of output manifest files for multi-label image classification labeling job, see [Multi-label classification job output](sms-data-output.md#sms-output-multi-label-classification).

# Image Label Verification
<a name="sms-label-verification"></a>

Building a highly accurate training dataset for your machine learning (ML) algorithm is an iterative process. Typically, you review and continuously adjust your labels until you are satisfied that they accurately represent the ground truth, or what is directly observable in the real world. You can use an Amazon SageMaker Ground Truth image label verification task to direct workers to review a dataset's labels and improve label accuracy. Workers can indicate if the existing labels are correct or rate label quality. They can also add comments to explain their reasoning. Amazon SageMaker Ground Truth supports label verification for [Classify image objects using a bounding box](sms-bounding-box.md) and [Identify image contents using semantic segmentation](sms-semantic-segmentation.md) labels. You create an image label verification labeling job using the Ground Truth section of the Amazon SageMaker AI console or the [CreateLabelingJob](https://docs.aws.amazon.com/sagemaker/latest/APIReference/API_CreateLabelingJob.html) operation. 

Ground Truth provides a worker console similar to the following for labeling tasks. When you create the labeling job with the console, you can modify the images and content that are shown. To learn how to create a labeling job using the Ground Truth console, see [Create a Labeling Job (Console)](sms-create-labeling-job-console.md).

![\[Example worker console for labeling tasks, provided by Ground Truth.\]](http://docs.aws.amazon.com/sagemaker/latest/dg/images/label-verification-example.png)


You can create a label verification labeling job using the SageMaker AI console or API. To learn how to create a labeling job using the Ground Truth API operation `CreateLabelingJob`, see [Create a Labeling Job (API)](sms-create-labeling-job-api.md).