Most Hollywood thriller movies reach a pivotal point where the cops scan through their criminal records trying to match images or computerized face scans with a prospective criminal. This is usually followed by an exact match highlighted on the screen and guns being pulled out of holsters with a sudden urgency to take action. If you do recollect such scenes, then you have witnessed an excessive dramatization of the Facial Recognition technology.
Facial Recognition today has come far beyond just recognizing criminal faces and barged into our daily lives.
Facial Recognition is a ground-breaking technology; an AI-based biometric software simply built on a deep learning algorithm. It is highly used in unlocking smartphones, preventing retail crimes, finding missing persons, helping the blind, protecting law enforcement, identifying people on social media platforms, diagnosing diseases, facilitating secure transactions, etc. In addition, it is used widely in getting authorized access to sensitive data or restricted areas.
Some of the noted Facial Recognition apps are Google Vision API, Amazon Rekognition, Kairos, OpenCV, Microsoft Face API (AKA ‘Project Oxford’), IBM Watson Visual Recognition API, NEC Face Recognition, Affectiva, etc.
There are various case studies conducted to test the accuracy of these different facial recognition technologies and one of the interesting ones was the case study conducted on the characters of the sensational Marvel movie, The Black Panther. Some applications failed in recognizing age, some failed in detecting genders, and some failed in detecting the races of the actors.
With a strong API integration system, AWS Rekognition is one of the leading face recognition applications with accurate face, object, and scene detection with identity and access management. Amazon Rekognition is extensively used for image and video analysis in applications. You just provide an image or video to the Rekognition API, and the service can identify the objects, people, text, scenes, and activities, as well as detect any inappropriate content.
Features of AWS Rekognition
Let’s discuss some of the striking features of AWS Rekognition:
Object, Scene, and Activity Detection
With Amazon Rekognition, you can identify thousands of objects (e.g. bike, telephone, building, etc.) and scenes (e.g. parking lot, beach, city, etc.). When analyzing video, you can also identify specific activities occurring in the frame, such as "delivering a package" or "playing soccer".
Facial recognition
Rekognition’s fast and accurate search capability allows you to identify a person in a photo or video by using your private repository of face images.
Facial analysis
You can analyze the attributes of faces in images and videos you provide to determine things like emotions, age range, eyes, glasses, facial hair, etc. In the video, you can also measure how these things change over time, such as building a timeline of the emotions of an actor.
Pathing
You can capture the path of people in the scene by using Amazon Rekognition with video files. For example, you can use the movement of athletes during a game to identify the playing tactics for post-game analysis.
Unsafe Content Detection
Amazon Rekognition helps you identify potentially unsafe or inappropriate content across both image and video assets. It also offers detailed labels that allow you to accurately control what you want to permit as per your needs.
Celebrity Recognition
You can quickly identify well-known people in your video and image libraries to enlist footage and photos for marketing, advertising, and media usages.
Text in Images
Specifically built to work with real-world images, Rekognition can detect and recognize text from images, such as street names, captions, product names, and license plates.
Some of the other features are:
- Searchable image library
- Image Moderation
- Face based user verification
- Sentiment Analysis
- Facial Recognition
- Celebrity Recognition
Facial Recognition with Amazon Rekognition
Amazon Rekognition makes it easy to search your image collection for similar faces by storing face metadata using the IndexFaces API function. You can use the SearchFaces function to return high confidence matches. A face collection is an index of faces that you own and manage.
Below are the steps to give a brief overview of facial recognition:
- Adding faces to pull: we need to add the faces in recognition pull(collection), using IndexFaces API function. Images are passed to collection as base 64 encoded image or reference to an image from S3 bucket. Image format must be .png or .jpeg.
Ex. var params = {
CollectionId: "images-collection", //rekognition collection name
ExternalImageId:”Image_id”, //name to be given for that face in collection
Image: {
S3Object: {
Bucket: "img-recognition-pull",
Name: “Image_name”
}
}
};
rekognition.indexFaces(params, function (err, data) {
…..
…..
});
- Recognizing input face against pull: The face to be recognized can be provided as input through an image from S3 bucket or base 64 encoded image. Input image should be .jpeg or .png. The largest face from that image is considered to be the face to be recognized.
Ex:
var params = {
CollectionId: "images-collection", //mane of rekognition pull
FaceMatchThreshold: 75, //minimum percentage to consider face matching
Image: {
S3Object: {
Bucket: "img-recognition-pull", //source image S3 bucket name
Name: “image-name” //name of source image
}
},
MaxFaces: 5 //maximum number of faces to match from collection
};
rekognition.searchFacesByImage(params, function (err, data) {
…..
…..
});
Moving On
With the inception of facial recognition technologies, we have advanced miles ahead in the personalization of services. Services like AWS Rekognition are giving us a leverage in areas like security management, Health, Marketing and Social Media, and much more. Soon we will see a surge in hybrid technologies in combination with Face Recognition, like Biometric and geolocation, catering to our day-to-day needs. The sooner we start incorporating them in our solutions the closer we get to our customers.
