Content Moderation:User guide for the Image Moderation 2.0 console

When you first use the service, you can log on to the Rules page in the console to view the default detection scope configurations.

Scenarios

If your business has the following requirements, you can use the Alibaba Cloud Image Moderation V2.0 console to customize detection rules and scopes. You can also use the console to query detection results and view usage statistics.

Prerequisites

Adjust the risk detection scope for images

You can adjust the detection scopes and specific items for risk detection based on common industry practices or your business needs.

1. Log on to the Content Moderation console.

2. In the navigation pane on the left, choose Machine Moderation V2.0 > Image Moderation > Rules.

3. On the Rules Management tab, find the service that you want to manage. This example uses the Common Baseline Moderation (baselineCheck_global) service. Click Settings in the Actions column.

4. On the Detection Scope page, select a detection type to adjust. This example uses Prohibited Content Detection.

   1. On the Prohibited Content Detection tab, view the default configurations in the Detection Scope Configuration section. For example, the following figure shows that four items are detected by default. If an item is hit, the corresponding label is returned.

      

   2. Click Edit to enter edit mode. Change the Detection Status as needed. For example, the following figure shows that the detection switch for the third item is turned off.

      

   3. You can also adjust the Medium-risk Score and High-risk Score for labels to determine the returned risk level.

      Note

      Risk level rules:

      1. If a risk label is detected and the confidence score is within the high-risk score range, the risk level is high.

      2. If a risk label is detected and the confidence score is within the medium-risk score range, the risk level is medium.

      3. If a risk label is detected and the confidence score is below the medium-risk score range, the risk level is low.

      4. If multiple labels with different risk levels are hit, the highest risk level is returned.

      5. If no risk labels are hit, the risk level is safe.

      6. If a custom blocklist is hit, the risk level is high.

   4. Click Save to save the new detection scope. The new configuration takes effect and is applied to the production environment in about 2 to 5 minutes.

Set different risk detection scopes for multiple business scenarios

You can copy a service and set different risk detection scopes to meet the different moderation requirements of your business scenarios.

1. Log on to the Content Moderation console.

2. In the navigation pane on the left, choose Machine Moderation V2.0 > Image Moderation > Rules.

3. On the Rules Management tab, find the service that you want to copy. This example uses the Common Baseline Moderation (baselineCheck_global) service.

   1. On the service list page, click Copy in the Actions column for the Common Baseline Moderation service.

   2. In the Copy Service panel, enter a Service Name and Service Description.

      

   3. Click Create to save the copied service information. The new service can be called 1 to 2 minutes after it is created.

   4. After the service is created, you can configure and edit the copied baselineCheck_global_01 service. You can call the baselineCheck_global_01 service and the baselineCheck_global service separately to meet the needs of different business scenarios that require different risk detection scopes.

Perform special or emergency moderation on specific known images

You can configure a custom image library for images that may have risks. If a user-uploaded image matches an image in the configured library, a risk label is returned.

1. Log on to the Content Moderation console.

2. Before you configure a custom image library, you must create and maintain the library. If an existing library meets your business requirements, you can skip this step.

   Note

   Each account can create up to 10 image libraries. The total number of images in all libraries cannot exceed 100,000.

   • Create an image library and upload images

     1. In the navigation pane on the left, choose Machine Moderation V2.0 > Image Moderation > Image Libraries.

     2. Click Create Image Library. In the Create Image Library dialog box, enter a library name and description, and then click OK.

     3. Find the library that you created and click Image Detail in the Actions column.

     4. Click Add Image. In the Add Image dialog box, Select Image and follow the instructions to upload images.

        You can upload a maximum of 10 images at a time. Each image must be 4 MB or smaller. We recommend that the image resolution be at least 256 × 256 pixels. The upload list shows the upload status of up to 10 images. To upload more images, clear list and continue uploading.

        

     5. On the details page of the image library, you can view the list of uploaded images. You can also query and delete images.

        • Query images: You can search for images by Image ID or Add Time.

        • Delete images: You can delete images from the library, including in batches.

        

   • Maintain an existing image library

     1. In the navigation pane on the left, choose Machine Moderation V2.0 > Image Moderation > Image Libraries.

     2. Find the library that you want to maintain. Click Edit in the Actions column to modify the library name and description. Click Image Detail in the Actions column to upload or delete images.

3. In the navigation pane on the left, choose Machine Moderation V2.0 > Image Moderation > Rules.

4. On the Rules Management tab, find the service that you want to manage. This example uses the Common Baseline Moderation (baselineCheck_global) service. Click Settings in the Actions column.

5. On the Detection Scope page, select a detection type to adjust. This example uses Prohibited Content Detection.

   1. In the Set Labels by Customized Libraries section of the Prohibited Content Detection tab, view the information about the configured custom image library.

      

   2. Click Edit to enter edit mode and select the custom image library that you want to configure.

      

   3. Click Save to save the new custom image library configuration.

      The new configuration takes effect and is applied to the production environment in about 2 to 5 minutes. If a user-uploaded image matches an image in the configured library, the "contraband_drug_lib" label is returned.

Exempt trusted images from risk detection

You can exempt trusted image libraries from risk detection to prevent images from being flagged as risky by Content Moderation's algorithms.

1. Log on to the Content Moderation console.

2. In the navigation pane on the left, choose Machine Moderation V2.0 > Image Moderation > Rules.

3. On the Rules Management tab, find the service that you want to manage and click Settings in the Actions column.

4. Click the Exemption Configuration tab and modify the configuration.

   1. On the Exemption Configuration tab, view the list of custom image libraries and their exemption status.

      For example, the following figure shows that the exemption switches for all image libraries are turned off.

   2. Click Edit and turn on the exemption switch for the image library that you require.

      

   3. Click Save to save the new configuration for exempted images.

      The exemption library configuration takes effect in about 2 to 5 minutes. Alibaba Cloud Content Moderation compares the input image with each image in the selected libraries for similarity. For images that the algorithm considers identical, the Content Moderation system returns the "nonLabel_lib" label and no other risk labels.

Customize detection settings for text in images

Test image moderation effects online

View statistics on recent image detections

You can view statistics on the volume of recent image detections to formulate further moderation or governance measures for specific image content.

1. Log on to the Content Moderation console.

2. In the navigation pane on the left, choose Machine Moderation V2.0 > Image Moderation > Usage Statistics.

3. On the Usage Statistics page, select a time range to query or export usage data.

   • Query Usage: You can view usage statistics by day or by month. The statistical data is stored for one year. You can query data spanning up to two months.

   • Export Usage: Click the icon in the upper-right corner to export usage data by day or by month.

     The exported report is in Excel format and includes only data with call volumes. The following table describes the row headers in the exported report.

     Name	Description	Unit
     Account UID	The UID of the account that exported the data.	None
     Service	Information about the called detection service.	None
     Usage	The total number of calls.	Count
     Date	The date the statistics were collected.	Day/Month

   • View service hit details: After the usage statistics are generated, the label hit details for each called service are displayed. The details are shown as a column chart of daily call volumes and a treemap chart of label proportions.

     • Column chart of daily call volumes: Shows the number of daily requests that hit risk labels and the number of requests that did not hit risk labels.

     • Treemap chart of label proportions: Shows the overall label hit situation, arranged in descending order of label proportion. Labels with the same prefix have the same background color.