This repository demonstrates the results of an AWS Face Recognition Service built using AWS Rekognition. The project uses a serverless architecture with Amazon S3, AWS Lambda, and DynamoDB to provide a scalable and efficient face recognition solution.
- Serverless Architecture: Event-driven design using AWS Lambda.
- Facial Recognition: Efficient face matching with AWS Rekognition.
- Image Storage: Centralized storage of images in Amazon S3.
- Data Management: Metadata and face data stored in DynamoDB.
- Real-Time Processing: Automatic processing of images upon upload.
Here are the results from the implemented AWS Face Recognition Service:
-
Name: Bill Gates
-
Confidence: 99.98%
-
Face ID:
0c38467a-f79d-450d-9823-76616f1c825c -
Name: Elon Musk
-
Confidence: 99.97%
-
Face ID:
a2292089-9f6c-4f27-ac60-e23df24c40a5 -
Name: Sundar Pichai
-
Confidence: 99.87%
-
Face ID:
5fa0066-7d60-415b-a007-3ac75b582e95
An unrecognized individual's face, which isn't stored in DynamoDB, will be flagged as "No Match."
For a step-by-step guide to implementing this project, including the setup of S3, Lambda, Rekognition, and DynamoDB, visit the detailed instructions here:
The system follows this workflow:
- Images are uploaded to an S3 bucket.
- A Lambda function triggers automatically to process the uploaded image.
- AWS Rekognition extracts and analyzes facial data.
- The processed data is stored in DynamoDB for future recognition.
- If a match is found, the name and confidence score are returned. If no match exists, it is flagged as unrecognized.
This project is licensed under the Apache License 2.0. See the LICENSE file for details.