
Welcome to this course on Amazon Rekognition
Introduction to Deep Learning, Amazon Rekognition Service and Demo using the AWS Management Console
You will need an AWS free tier account, A computer with a Browser, Node.js and a Code Editor installed
Getting Started with Amazon Rekognition
Build a web page to upload a customer image to an S3 bucket and display resulting ad image via an API Gateway backed service, using Cognito unauthenticated identities for access.
Create a lambda function named indexface that triggers on uploads to the face index bucket to index faces with Amazon Rekognition, creating Friends Family collection if needed and logging results.
Create another S3 bucket to act as an input repository in us east region, set a bucket policy allowing anonymous read and write from any domain, and save.
Create a lambda function that detects all faces, crops each, and searches a Rekognition collection, using async waterfall and parallel processing to overcome the largest-face limitation.
Expose a rest API named picture album with Amazon API Gateway integrated to a Lambda function, configure filename query parameter and body mapping templates, test, and deploy to Dev.
Use the enhanced web client to upload and index facial images, run searches against the collection, and review results and logs via CloudWatch and Lambda.
Set up a Cognito identity pool, create a web client, and test verification by uploading an image that triggers a lambda call to compare with a reference.
Explore how the Detect Labels API detects objects, scenes, and concepts from images, with S3 or direct input, and index results in ElasticSearch for fast search.
Create an elasticsearch domain in the management console, select the latest 5.3 engine and smallest instance, enable open access, then review and create to obtain the endpoint for a lambda.
Walks through creating a lambda function that detects image labels with Rekognition and indexes them into Elasticsearch when images are uploaded to the travel history bucket.
Expose a travel api with Amazon api gateway and connect it to a lambda function, configure request mapping and a search query string, enable cors, and deploy.
Build a web client to upload travel photos and search them using a travel API, wired with identity pool credentials and an API gateway, with index.html, cascading stylesheet, and JavaScript.
Learn to use the DetectModerationLabels API to detect explicit and suggestive content, understand the parent-child label hierarchy with confidence scores, and implement browser-based moderation via S3 or base64 input.
Explore how the RecognizeCelebrities API identifies famous individuals from an uploaded image, returning matched celebrity names, confidence scores, bounding boxes, landmarks, and optional links for more info.
Create a web client that lets users select an image, calls recognised celebrities api to detect faces, and displays each celebrity name in album list while testing with two images.
Amazon Rekognition is a sophisticated deep learning based service from Amazon Web Services (AWS) that makes it easy to add powerful visual search and discovery to your own applications. With Rekognition using simple APIs, you can quickly detect objects, scenes, faces, celebrities and inappropriate content within images. You can also search, analyze, organize and compare faces.
"I love how you structured the course (demo and code) and most importantly you showed how to use the technology. Most of the course that I've seen and taken lack the latter part." - Sovanna Chuon, Student
"It is exactly what I was looking for: a quick and clear introduction to AWS Rekognition." - Miguel Cabral, Student
"Buen curso, ritmo adecuado y entendible, tal vez falta entrar más en detalle en el código pero en general muy recomendable." - Eduardo Vargas Aguirre, Student
Are you a Developer looking to incorporate visual search and Image recognition capabilities to your own Applications using the recently launched Amazon Rekognition service from AWS? Are you an Architect or Tech Lead willing to explore reference Architectures and implementations on how Amazon Rekognition fits in with other AWS service offerings to solve Real-world problems? Either way, you are in the right place.
This course emphasize on learning by doing. You will be developing 6 complete end-to-end solutions with different capabilities of Amazon Rekognition. You will be developing the following projects:
Deliver Dynamic Targeted Advertisement Content
Intelligent Family Photo Search App
Face Based Authentication and Verification System
Travel Photo Smart Search App
Enhance the Travel Photo Smart Search App with Image Moderation
Celebrity Recognizer
In the process, you will also interact with and learn to use the following AWS services
S3
Lambda Function
API Gateway
DynamoDB
ElasticSearch
Cognito
SNS
CloudWatch
Identity and Access Management
***Important Update - ImageMagick Fix***
September 15, 2019
ImageMagick is no longer bundled with the Node.js 10.x runtime in Lambda due to security reasons. I have included a lecture with all the instructions to install and use a lambda layer to address this issue.