Amazon Rekognition

Amazon Rekognition deep learning software makes it easy to add video and image analysis to applications.

Overview

Amazon Rekognition allows users to add video and image analysis to their applications easily. Users provide a video or image to the Amazon Rekognition API, then the service would, in turn, identify activities, scenes, text, people, objects, as well as identify any inappropriate content. The software also offers highly accurate facial search, facial comparison, and facial analysis capabilities. Newspaper and non-profit organizations can also compare, analyze, and detect face for multiple varieties of public safety, people counting, user verification, and cataloging use cases.

The software is a highly scalable, deep learning technology that allows users to analyze millions of videos and images every day. Users do not need to have any knowledge of machine learning to use Amazon Rekognition software. Also, the software includes an easy-to-use that analyzes any video and image file that users store in Amazon S3.

Amazon Rekognition has custom labels that users can use in identifying scenes and objects in images that fit their business requirements. Users can also develop a model that classifies machine parts or unhealthy plants on their assembly lines.

The software also has Customs Labels that takes care of the technicalities of model development. Users only have to supply the software with the images of scenes or objects that they want to identify, and the system produces appropriate results.

Product Details

Amazon Rekognition deep learning software offers searchable video and image libraries. The software makes it possible for users to search through stored videos and images. This way, users can identify scenes and objects within those videos and images.

Amazon Rekognition deep learning software supports facial user verification. Users can configure their applications to check user identities by matching a reference image with a live image.

Amazon Rekognition deep learning software offers users demographic and sentiment analysis. The software helps users to interpret emotional expressions like a surprise, sadness, or happiness. It also interprets demographic data like someone's gender from facial images. The software allows users to analyze images, and send the demographic attributes and emotions to Amazon Redshift. Redshift then creates periodic reports on trends like in-store locations and other related scenarios. The software predicts emotional expressions using only the physical appearance of a person's face.

Amazon Rekognition deep learning software supports facial searches. Users can search streaming videos, stored videos, and images for faces that match those stored in a "face collection." The face collection is an array of faces managed and owned by users. The software searches for a person's face by indexing the face and searching the user's face collection.

Amazon Rekognition deep learning software allows users to detect unsafe content. It can detect violent and adult content in stored videos and images. Also, developers can use the reported metadata to screen inappropriate content, depending on their specifications.

The API flags images that contain unsafe content, and also produces a ranked list of labels and their confidence scores. The labels specify the categories of unsafe content, and this enables the granular management and filtering of substantial volumes of user-generated content. This feature is especially useful for dating sites, online advertising services, entertainment, e-commerce sites, children apps, forums and blogs, and photo-sharing platforms. Also, users can utilize detailed labels and confidence scores to set up different business rules.

Amazon Rekognition deep learning software supports celebrity recognition. The software recognizes celebrities depending on the videos and images supplied by users. Also, Amazon Rekognition can identify tons of celebrities across different categories like media, entertainment, business, sports, and politics.

Amazon Rekognition deep learning software offers text detection. The API's Text-in-Image feature allows users to identify and extract content from their images. It also supports multiple fonts, even stylized ones. The software detects numbers and texts in different orientations like those found in posters and banners. Users can also use the software in social media and image sharing applications to enable searches based on a collection of images containing the same keywords.

Also, it can be used in entertainment and media applications to index videos based on specific texts on screens like captions, sports scores, news, and ads. This feature can also be used in public safety applications to identify cars using the license plates numbers from pictures taken by street cameras.

Amazon Rekognition deep learning software also supports custom labels. Users can use this feature in identifying scenes and objects in images that fit into their requirements. They can identify animated characters in videos, tell the difference between infected and healthy plants, group machine parts in their assembly lines, identify their product on store shelves, and find their logos in social media posts.

Amazon Rekognition deep learning software simplifies data labeling. With this feature, users have access to a visual interface that makes imaging labeling quick and easy. Also, users can label and identify specific objects in images with bounding boxes or label entire images, all with an easy-to-use click-and-drag interface.

Amazon Rekognition deep learning software supports person pathing. Users can use this feature in capturing how, when, and where each person moves in their video. The software also offers a specific index for each person found in videos. This way, users can identify the number of people in their videos.

Amazon Rekognition deep learning software offers users administrative security. The API is integrated with AWS Identity and Access Management. These policies allow users to control access to the API and also oversee resource-level permissions for their accounts.

Recap

Amazon Rekognition deep learning software enables developers to add video and image analysis to their applications. A core strength of the Amazon Rekognition deep learning software is that developers do not need to have prior knowledge of machine learning and deep learning technologies to take advantage of the software. This API can identify any inappropriate content, activities, scenes, text, people, and objects from a video or image. The software also provides users with fast and accurate face searches. This way, they can identify a person from a video or photo using a repository of face images or images that they have stored for comparison.