Amazon Rekognition

Amazon Rekognition is an image recognition service that detects objects, scenes, faces, text, and inappropriate content in images. You can also use the service to search and compare faces.

Introduction to the Amazon Rekognition Activities Package

The Amazon Rekognition Activities Package extends the capabilities of UiPath Studio by integrating the robust Amazon Rekognition image service. Using the Amazon Rekognition Image APIs, this activities package provides you with the following image analysis capabilities:

1. Face Detection, Identification, and Comparison
2. Distinct Object Identification (Label Detection)
3. Unsafe Content Detection
4. Text Detection
5. Creation of Face Collections

Integrating Amazon Rekognition with UiPath

The Amazon Rekognition Activities Package facilitates seamless interaction between UiPath and Amazon Rekognition by establishing an authenticated connection to your AWS account through the Amazon Scope activity. Once this connection is established, you can leverage the other activities within the package to automate various tasks involving the Amazon Rekognition Image service.

Using Amazon Rekognition Activities

Each Amazon Rekognition activity invokes a specific Amazon Rekognition Image API action, utilizing user-defined input properties (i.e., request parameters). Upon successful execution, the activity delivers relevant response elements (i.e., output properties) that can be used as inputs in subsequent activities, queued for further processing, and more.

Preparing for Project Development

Before you build your first project, complete the steps in the Setup guide.

Following the setup, the Quickstart guides offer step-by-step instructions to create samples of the activities to validate the connection to your AWS account and facilitate familiarity with the input/output properties.

Activities in the Amazon Rekognition Activities Package

The Amazon Rekognition Activities Package encompasses the following activities:

Amazon Scope

Establishes an authenticated connection between UiPath and your AWS Account using your IAM credentials.

Add Face

Detects faces in the input image and subsequently adds them to the specified collection. For each detected face, the activity furnishes an array of the added faces.

Analyze Face

Identifies the 100 largest faces in the input image and provides an array of face detail objects for each detected face.

Compare Faces

Compares the most prominent face in the source input image with each of the 100 largest faces detected in the target input image.

Create Collection

Stores information about detected faces in a server-side container (i.e. collection). You can use the facial information that's stored in a collection to search for known faces in images using the Identify Face activity.

Identify Face

Initially detects the largest face in the input image, then searches the specified collection for matching faces. The activity assesses the features of the input face and outputs an array of faces that match the input image.

Detect Labels

Identifies instances of real-world objects within an input image, including items like flowers, trees, and tables; events such as weddings, graduations, and birthday parties; and concepts like landscapes, evenings, and nature.

Detect Text

Detects the words, lines, and text found in the input image.

Safe Search

Detects unsafe content such as explicit nudity and suggestive material within your input image.

Applications of the Amazon Rekognition Activities Package

The Amazon Rekognition Activities Package can be of help in a wide array of use cases:

• Security and Access Control - Face Detection and Identification can be used to enhance security systems by recognizing authorized personnel for access control to restricted areas.
• Content Moderation - Automatically flag or filter out inappropriate content in user-generated images or uploaded media on websites and social platforms, ensuring a safe and user-friendly environment.
• Medical Imaging - Automation can help in the healthcare industry for medical image analysis, such as detecting specific patterns in X-rays, MRI scans, or identifying specific medical equipment in images.