AWS re:Invent 2017: Media Intelligence for the Cloud with Amazon AI (MAE402)

Опубликовано 28 ноября 2017, 18:42

In this session, we take a pragmatic approach to enhancing common media workflows built around ingest, media asset management, live video, and OTT on-demand streaming. We show how to extract metadata as an additional intelligence layer for video using Amazon AI services, such as Amazon Rekognition, in combination with turnkey architecture built around AWS Lambda, Amazon ECS, and Amazon EC2 Spot Instances. The capabilities offered by Amazon AI services provide a unique opportunity to eliminate the traditional undifferentiated heavy lifting associated with contextual, facial recognition and object-based media metadata creation—that is, who is in with what, and where. We also discuss a large studio and broadcaster just starting to use these intelligent offerings from AWS as they change their method of how to best leverage the business value of their content.