Meta Leadership Primer: Amazon Rekognition

Image recognition is not anything new. After all, if you have an account on one of many social media platforms, you probably already used it to tag one of your friends. However, this technology has been going to the gym and spending time at the library (showing my age, I know) lately! Getting bigger, faster and smarter!

One of the most recent upgrades of image recognition can be seen on the Microsoft Surface Book 2 where you can use facial recognition for device authentication. Besides, there are also some high-end cameras on the market that allow you to immediately recognize the person as well as they are even able to identify if the person is smiling.

We have all heard the cliché “a picture is worth a thousand words”. And that is precisely why this imaging technology has been evolving in the last few years. Guess what? This is only the beginning stages and the industry is in my opinion and the last stages of its infancy with some substantial growth set for its future.

Recently, Amazon started offering its own image and video recognition service — the Amazon Rekognition. The Amazon Rekognition uses the Amazon Rekognition API.

I have actually tested this out from an API offered in the book: AWS Lambda in Action: Event-driven serverless applications by Danilo Poccia. In which there are actually code snippets you can use for testing and proof of concept establishment. Its actually pretty robust in that it can detect faces, count the number of people and even render a opaque cover of images that are deemed sexually suggestive. Pretty nice for image analysis on the fly. Granted, this was in 2017 and that technology has gotten considerable better by now, with the addition of Video. Wait, What? Yup, the service can extrapolate images from video for analysis!!!!!

Learn more about image analysis types of technologies and other “Next Practices” in Meta Leadership

Visit Michael Stattelman for even more….


What Is Amazon Rekognition?

Simply put, when you leverage Amazon Rekognition, you will be able to easily add image and video analysis to your applications. All you need to do is to pipe the image or video through the Rekognition API and it will immediately identify activities, scenes, text, people, and objects.

One of the things that I should mention is that the Amazon Rekognition delivers a highly accurate facial recognition and facial analysis. So, you will be able to easily compare, analyze, and detect faces. You can use this for a wide variety of situations including public safety, people counting, cataloging, and user verification.

Benefits Of Amazon Rekognition

#1: Deep Learning-Based Video And Image Analysis:

The Amazon Rekognition uses a deep learning technology that allows it to accurately analyze images, compare and find faces in images, as well as detect scenes and objects on both videos and images. Much faster than the human equivalent.

#2: Integrate Image & Video ReKognition Into Your Apps:

One of the best things about using Amazon Rekognition is that you don’t need any deep learning expertise or computer vision. The reality is that you won’t have any problems integrating the Rekognition API into any mobile, web or connected device application. As long as all of the params are passed along in the expected format and you have a return that catches the output of the system.

#3: Scalable Image Analysis:

In case you need to deal with a lot of images, you can definitely use Amazon Rekognition to save you time. After all, with the Amazon Rekognition, you will be able to analyze millions of images that will allow you to organize and curate huge visual data. Please understand that this scalability comes at a cost and while the AWS free Tier is enough for most small businesses and use cases for proof of concept generation.

#4: Low Cost:

Just like all Amazon products, the Amazon Rekognition is extremely affordable and perfect for the little guy. After all, there are no minimum fees or upfront commitment and you only pay for what you have there. Once you start scaling up to massive amounts of image or video data it can get really costly, really quick! How fast you might say. I was made aware of a situation in a former organization in which the spawning of virtual machines was set at a really low threshold and the alert sent out for spending limits of AWS services hit $30k in 10 minutes! That’s right kids, THIRTY THOUSAND USD in 10 minutes. Just ensure that you have knowledgeable people when testing and deploying cloud services.

#5: Easy To Integrate With Other AWS Services:

In case you decide to use the Amazon Rekognition, one of the great features is the complete integration with other AWS services such as the AWS Lambda or the Amazon S3. This allows a complete build if you so choose on one provider system if you so choose.

#metaleadership (Michael Stattelman)

Written by

Meta Leadership #metaleadership @themetaleadership