Amazon Rekognition
Amazon Rekognition
1h 11m

In this lecture we'll now start diving into the Amazon Rekognition service and how it works. We discuss what motivations may exist that could require you to consider using the Amazon Rekognition service. We provide a brief understanding of how Rekognition uses machine learning technology under the hood. We'll discuss how the Amazon Rekognition API is structured into operations specific to image processing and those that are specific to video processing.


Welcome back. In this lecture, we'll now start diving into the Amazon Rekognition Service and how it works. Okay, let's begin. Amazon Rekognition is a service that enables you to quickly and easily integrate computer vision features directly into your own applications. At its core, the Rekognition service provides an API that you submit images and or videos to. You then instruct the Rekognition service to perform a specific analysis on the media. The analysis can be anything from detecting faces with an image to extracting labels or features from a video in a synchronized manner.

Under the hood, Amazon Rekognition uses deep learning technology. The service employs a convolution neural network, CNN, having been pre-trained on a massive amount of labeled ground truth data. Amazon takes ownership to build and train the underlying models, which are both time-consuming and computationally expensive tasks. As such, the end result enables you to leverage the computer vision features that Amazon Rekognition provides, without having to invest time into building your own deep learning models, nor having to consider or understand the complexities involved in operating a GPU-enabled cluster to build them, and the associated expense as well.

Amazon Rekognition is broadly separated into operations specific to processing images and operations specific for processing videos or video streams. The main difference between these two sets of APIs is that the image processing APIs follow a synchronous pattern, whereas the video processing APIs follow an asynchronous pattern for calling.

In this course, we'll cover both sets of APIs in separate lectures. Additionally, the Rekognition API operations are grouped into those that are considered storage-based and those that aren't. The storage-based APIs support persisting certain facial features and server-side containers known as collections. We'll cover off collections in a separate lecture.

But for now, that concludes our lecture for the Intro into the Amazon Rekognition Service. In the next lecture, we'll focus on the Rekognition image processing APIs. Here we'll learn about the individual API operations and how we go about using them to perform image analysis. Go ahead and close this lecture and we'll see you shortly in the next one.

About the Author
Learning Paths

Jeremy is a Content Lead Architect and DevOps SME here at Cloud Academy where he specializes in developing DevOps technical training documentation.

He has a strong background in software engineering, and has been coding with various languages, frameworks, and systems for the past 25+ years. In recent times, Jeremy has been focused on DevOps, Cloud (AWS, Azure, GCP), Security, Kubernetes, and Machine Learning.

Jeremy holds professional certifications for AWS, Azure, GCP, Terraform, Kubernetes (CKA, CKAD, CKS).