The automated review service for VOD media processing allows you to moderate media assets by detecting pornography, terrorism and political content, text and image violations, logo recognition, undesirable scenes, and voice content violations. This topic describes each review type and how to use them.
Automated review types
Automated review accurately identifies risky or non-compliant content in video, audio, and image files. The automated review feature for VOD media processing currently supports the following six review types:
|
Type |
Description |
|
Pornography |
Detects pornographic or sexually suggestive content. |
|
Terrorism and political content |
Detects terrorism-related or politically sensitive content. |
|
Text and image violations |
Detects non-compliant text and images, including advertisements. |
|
Logo recognition |
Detects logos, such as TV station icons and trademarks. |
|
Undesirable scenes |
Detects undesirable scenes, such as black screens, pillarboxing or letterboxing, dim lighting, Picture-in-Picture (PiP), smoking, and in-car streaming. |
|
Voice content violations |
Detects voice spam, such as advertisements, political content, or terrorism-related speech. |
Create an automated review template
Console
API
Create an automated review job
Console
API
Query job details
Callback
API
Query job usage
-
Log on to the Intelligent Media Service console.
-
In the left-side navigation pane, choose Data Center > Usage.
-
On the VOD Tasks tab, in the Video AI section, you can query the duration usage of jobs or export the usage details.
NoteTo ensure data security, the download link for the exported file is valid for 30 minutes. If the link expires, refresh the page and try again.