At a training session, it was noted that image-related AWS AI services are often confused. Which TWO of the following correctly pair the use cases of 'image analysis' and 'document text extraction (OCR)' with the corresponding AWS services? (Choose TWO.)
Select two correct pairings of image analysis and document OCR services.
Amazon Rekognition = image and video analysis (object and face detection, etc.)
This is correct. Amazon Rekognition is a service that analyzes images and videos to detect objects, faces, and text, corresponding to the use case of 'image analysis.'
Amazon Textract = document text extraction (OCR) and field extraction
This is correct. Amazon Textract is a service that extracts text, tables, and form fields from document images, corresponding to the use case of 'document text extraction (OCR).'
Amazon Rekognition = document text extraction (OCR)
This is incorrect — the roles are reversed. Rekognition's primary use case is image and video analysis (object and face detection). Extracting text, tables, and fields from documents is the role of Textract. Watch out for this common image-related confusion.
Amazon Textract = detecting objects and faces in images
This is incorrect — the roles are reversed. Textract's primary use case is extracting text, tables, and form fields from document images (OCR). Detecting objects and faces is the role of Rekognition.
Amazon Rekognition = audio transcription
This is incorrect. Rekognition is a service for image and video analysis and cannot handle audio. Audio transcription is the role of Transcribe.
The correct service-to-use-case pairings are: 'Amazon Rekognition = image and video analysis (object, face, and text detection)' and 'Amazon Textract = document text extraction (OCR) and field extraction.' Options that swap the two roles are a classic trap. Also note that audio transcription is handled by Amazon Transcribe, not Rekognition.