New Technologies in Facebook

One of the biggest social networking websites, Facebook, is not only a hub for the community but has also become the point of interaction for brands and consumers. Not limiting itself out of the loop, Facebook launches its own range of tech-powered products.

 

Since using Facebook is free of cost but it allows its users to interact in a number of ways like creating your profile, a space to post your own content, joining or starting groups; it allows Facebook to have access to data related to human behavior, preferences, and demographics. The company earns most of its revenue from advertisements. But the process of aligning the right advertisement with the right audience is handled with Artificial Intelligence algorithms and machine learning.

 

In this blog, we will look at how Facebook is making use of technologies to manage its operations. 

 

(Related reading: How Facebook uses Big Data to enhance customer experience) 

 

Technology in Facebook

 

Let’s take a look at how Facebook stays ahead in the race.

 

By Using Machine Learning for Advertisements

 

Machine learning as an extension of Artificial Intelligence uses data and algorithms to imitate human understanding. Facebook uses machine learning to rank feeds, advertisements, search results, while also creating text understanding-based algorithms.

 

“New computer vision algorithms can “read” images and videos to the blind and display over 2 billion translated stories every day, speech recognition systems automatically caption the videos that play in your news feed, and we create new magical visual experiences such as turning panorama photos into fully interactive 360 photos.”

 

That is not all, with 1.9 billion daily active users and a consistent number of businesses vying for the attention of these users, Facebook uses machine learning to decide which advertisements should be shown to people based on two criteria: audience targeting and auction stage. 

 

Target audience is decided by advertisers while Facebook creates its own audience categories based on demographics, interests, behaviors, and the actions these users take while coming across an advertisement.

 

Ads enter the auction stage based on users, who have been selected as the target audiences by the advertisers. Facebook selects which ads to present to a person based on what an advertiser wants out of that interaction. For this process, they use machine learning, which provides them with an estimate or a prediction of any particular person’s prospect of taking the action stated by the advertisers. Data is collected on multiple factors like the type of ad, how are people interacting with it- hiding it, commenting, etc.

 

By Using AI for Facial Recognition and Smart Displays

 

Although Facial Recognition is not a new technology, recognizing a person based on their pictures is a level Facebook has achieved along with the accuracy of 97.35% on the Labeled Faces in the Wild (LFW) dataset that matches the approximate human-level performance.

 

Deep Learning is a tool of Artificial Intelligence which is a complex amalgamation of algorithms and neural networks which mimics the way humans gain knowledge. Facebook uses Deep Learning (DL) for Facial recognition and Textual analysis.

 

(Suggested Reading: Machine Learning vs Deep Learning: How are They Different?)

 

To recognize a face AI-based algorithm (DeepFace) goes through four parameters. First, the facial recognition technology analyses each image for its pixels and gradients and further classifies it according to matches regarding the shape or structure of a human face.

 

After detecting the face it aligns the picture against a 3D model of pre-existing data and further derives a representation of the face. At last, it classifies the face of the person. All of this is made possible with advanced algorithms-based AI.

---

Technologies that work behind Facebook’s latest inventions

---

However, Facebook does not stop here. Facebook Portal, a video communication device launched by them is a smart display device made to feel two people in the same room despite being on the other end of the video call. These range of devices are based on AI-powered smart camera and smart sound. 

 

The smart camera detects humans but not animals, and zooms or changes directions to keep the person in the frame while simultaneously not allowing pets to distract the camera. The smart sound is enabled by advanced microphones which recognize the primary sound and reduce all environmental distracting noises. 

 

Facebook also added some effects through the use of Augmented reality to increase the interaction on these video calls. This feature works in a similar fashion to Snapchat’s AR-powered lenses.

 

(Read to know more about Snapchat’s technology: How Snapchat uses technology for effective marketing)

 

Facebook also uses Deep Learning for textual analysis called DeepText to understand the text in terms of its context. DeepText uses networks to derive meaning behind words and their interchangeable meanings according to the words around them. This allows the algorithm to apply what it has learned in one language to another because it does not assign meanings to any words. 

 

(Watch this video to understand how DeepText works based on context.)

 

Social Communication in Virtual Reality

 

Facebook recently launched a Virtual Reality based application- Horizon Workrooms, a 3D Virtual office people can have access to with an Oculus VR headset. Or, log in through a standard microphone and camera but these will show up on a screen within the virtual workspace. Further, the app uses Augmented reality to present people as avatars with the ‘offices’.

 

"You can think about the metaverse as an embodied internet, where instead of just viewing content — you are in it," CEO Mark Zuckerberg told The Verge in an interview.

 

This expands the traditional scope of Virtual Reality and Augmented reality from consumer-based products towards communications technology.

 

In less than two decades, since its beginning in 2004, Facebook has changed the way of doing business by providing a direct link to consumers. With the introduction of a virtual reality workspace, if it succeeds, businesses with have to adjust yet again to the digital business landscape.