Serverless Image Catalog
Photo sharing websites often allow users to publish user-generated content such as text, images, or videos. Manual content review and categorization can be challenging. This solution enables the automation of these tasks. See the AWS Blog post Building a serverless image catalog with AWS Step Functions Workflow Studio for a detailed walkthrough.
The example implements the following workflow:
- An image stored in Amazon S3 is checked for inappropriate content using the Amazon Rekognition
DetectModerationLabels
API. - Based on the result of (1), appropriate images are forwarded to image processing while inappropriate ones trigger an email notification.
- Appropriate images undergo two processing steps in parallel: the detection of objects and text in the image via Amazon Rekognition’s
DetectLabels
andDetectText
APIs. The results of both processing steps are saved in an Amazon DynamoDB table. - An inappropriate image triggers an email notification for manual content moderation via the Amazon Simple Notification Service (SNS).
This repository contains two AWS Serverless Application Model (AWS SAM) templates:
starter/template.yml
intended as the starting point for following the walkthrough in the AWS Compute Blog post Building a serverless image catalog with AWS Step Functions Workflow Studio.solution/template.yml
for deploying the completed serverless image catalog solution.
Prerequisites
- An AWS account
- An AWS user with
AdministratorAccess
(see the instructions on the AWS Identity and Access Management (IAM) console) - AWS SAM CLI using the instructions here
Deploy using SAM
Step 1: Clone repository
git clone [email protected]:pbv/serverless-image-catalog.git
Step 2: Build the template
cd serverless-image-catalog/solution
sam build
Step 3: Deploy the application
sam deploy --guided
Security
See CONTRIBUTING for more information.
License
This library is licensed under the MIT-0 License. See the LICENSE file.