AWS Elemental Inference
AWS Elemental MediaLive includes features that are seamless implementations of the features of AWS Elemental Inference. Elemental Inference is an AI service that lets you easily apply machine learning foundational models to video, audio, and image content for automated analysis, classification, and insights generation.
MediaLive implements the following features of AWS Elemental Inference:
-
Smart crop, which lets you produce channel video outputs that are a different aspect ratio from the video source. For more information, see Smart crop video using Elemental Inference.
-
Event clipping, which lets you produce file clips from the channel video source. For more information, see Clipping video using AWS Elemental Inference.
Elemental Inference quotas
Elemental Inference has its own quotas. All of these quotas apply when you use the Elemental Inference features of MediaLive.
For example, there are Elemental Inference quotas relating to feeds and feed outputs. When you include Elemental Inference features in a MediaLive channel, feeds and feed outputs are created as follows:
-
One feed is created, regardless of the number of enabled features.
-
One output is created for each enabled feature in the channel.
Keep in mind that these feeds and outputs are part of your Elemental Inference quotas.
For information about the default values for quotas and which quotas can be changed (adjusted) see the Elemental Inference section in AWS General Reference.
For more information about changing Elemental Inference quotas, see the Service Quotas console
