Amazon Rekognition
A computer vision service for image and video analysis.
Exam Tip: Rekognition = image and video analysis. If the question involves analyzing images (faces, objects, moderation, labels), the answer is Rekognition.
Object and Scene Detection
- Identify objects, scenes, and activities in images and videos
- Detect thousands of object types (cars, furniture, animals, etc.)
- Confidence scores for each detection
Face Detection and Analysis
- Detect faces in images and videos
- Analyze facial attributes (age range, gender, emotions, glasses, beard)
- Face comparison (compare two faces for similarity)
- Face search (search faces against a collection)
- Celebrity recognition (identify famous people)
Custom Labels
- Train custom models to detect objects specific to your business
- No ML expertise required — provide labeled images, Rekognition trains the model
- Use Cases: Detect defects on manufacturing lines, identify custom products, specific brand logos
Content Moderation
- Detect inappropriate, unwanted, or offensive content in images and videos
- Categories: nudity, violence, drugs, offensive symbols, hate symbols
- Confidence thresholds for automated filtering
- Integration with Amazon Augmented AI (A2I) for human review of edge cases
Common Use Cases
- User-generated content moderation for social platforms
- Identity verification and face-based authentication
- Manufacturing quality control with custom labels
- Media asset management and searchable video libraries
- Public safety and surveillance analytics