ApsaraVideo VOD provides intelligent media processing and content generation features, such as automated review, media fingerprint, DataQ - Smart Tag Service, and smart thumbnail. These features recognize, analyze, and understand audio and video content to improve the efficiency and quality of content production. This topic describes the video AI features that ApsaraVideo VOD provides.
Introduction to video AI services
Alibaba Cloud video AI services can recognize, analyze, and understand audio and video content. You can use video AI services to:
Detect non-compliant video content.
Identify and search for duplicate or similar audio and video segments.
Recognize people, text, entities, scenes, and actions in videos.
Analyze and understand videos to generate video tags, recommend thumbnails, create GIFs, and produce video synopses.
Convert speech to text.
Video AI Features
Feature | Description | More information |
Automated review | The automated review service detects content such as pornography, suggestive material, terrorism, special attire, special logos, weapons, and political references in the video files, thumbnails, and titles of video-on-demand resources. The service then provides recommended actions. |
|
Media fingerprint | A media fingerprint uniquely marks a video, audio file, or image. It is stable and does not change with transformations such as format conversion, editing, splicing, compression, or rotation. The media fingerprint service extracts and compares fingerprint features from images and audio in videos. This helps find duplicate videos, trace the source of video segments, and identify original content. |
|
DataQ - Smart Tag Service | The DataQ - Smart Tag Service analyzes visual, text, speech, and behavioral information in videos. It uses multimodal information fusion and alignment to accurately recognize content. The service automatically generates multi-dimensional content labels for videos. This transforms unstructured information into a structured format. |
|
Smart thumbnail | The smart thumbnail service analyzes and understands video content. It extracts the five best screenshots to use as candidate thumbnails. It also extracts key frames from the video to automatically create an animated GIF thumbnail. |
|
Notes
Video AI features are available only in some regions. For more information about the supported regions, see Service regions.
All video AI features use the Video AI Processing Complete switch to send callback notifications for completed AI jobs. For more information about how to configure event notifications, see Callback settings.
Fees are charged for video AI features. For more information about billing, see Video AI billing.
Video AI processing can be performed only on audio and video files that are uploaded to ApsaraVideo VOD. For more information about how to upload media files, see Upload media.
Video AI processing flow
You can call an API operation to start a video AI job. You can retrieve the results in three ways. The following figure shows the processing flow.
