Posted On: May 31, 2018

The Media Analysis Solution is a reference implementation that helps customers process, analyze, and extract meaningful data from their audio, image, and video files. The solution combines Amazon Rekognition, to provide highly accurate object, scene and activity detection; facial analysis and recognition; and celebrity detection in videos and images, Amazon Transcribe, an automatic speech recognition service, and Amazon Comprehend, which provides automatic transcription of audio files and extraction of key phrases and entities from transcripts, to quickly and seamlessly obtain key details from their media files in their AWS accounts without machine learning expertise. The solution also includes a web-based user interface that customers can use to upload and search their media files in their AWS accounts.

The solution deploys Amazon S3 to store metadata and host the user interface; AWS Lambda to trigger the analysis and metadata services; Amazon Rekognition for image and video analysis; Amazon Transcribe to add speech-to-text capabilities; Amazon Comprehend to extract key phrases and entities; Amazon API Gateway for searching capabilities;  AWS Cognito for user authentication; and Amazon Elasticsearch Service cluster to store results. To learn more about the Media Analysis Solution, see the solution webpage.

The AWS Solutions team communicates AWS architectural best practices and develops standardized, automated solutions for the platform. Our offerings currently live on the AWS Answers webpage, where customers can browse common questions by category to find answers in the form of succinct Solution Briefs or comprehensive Solutions, which are AWS-vetted, automated, turnkey reference implementations that address specific business needs.