Amazon Rekognition
Image and video analysis
A ready-to-use computer vision service: detects objects, faces, text in images, inappropriate content, and celebrities in images/videos; Custom Labels lets you train your own labels.
- Cost$ Low
- Difficulty●●●●●
- Costs money when idle?No
- Serverless
- In examsCLF · SAA
Think of it asLike a security guard with a superb memory: one look at a photo and they know what's in it and who's there.
When to use it
✓ Use it when
You need to analyze images/videos without training your own vision models.
✗ Avoid it when
You need to extract text and tables from scanned documents → Textract.
Common uses
- Moderate user-uploaded images
- Face verification for eKYC
- Automatic tagging for photo libraries
What does it cost?
- Pricing model: Charged by number of images analyzed and minutes of video; storing face metadata and Custom Labels cost extra.
⚠️ Cost traps
Custom Labels models are billed hourly while running, even with no calls.
