The social media giant, Facebook, made several announcements regarding significant changes toward how Artificial Intelligence (AI), and the devices that use as such; can serve us on a daily basis.
Facebook released to the public a real-world sound simulator that researchers may use to train AI systems in virtual three-dimensional spaces with sounds that mimic those that occur indoors; opening up the possibility that an AI assistant may one day help you track down a smartphone ringing in a distant room — specifically this Facebook AI that is now in the works.
Read: Extra Layer of Security Through Identity Confirmation For Your Facebook Account
Facebook AI Intelligence
Dhruv Batra and Kristen Grauman, research scientists of Facebook told us about their rapid progress in training the Facebook AI as well. Moreover, they also said that they are looking to complete a series of simulators to help in the process.
We are making rapid progress in training [AI] agents to accomplish a wide range of challenging tasks in photorealistic 3D simulators featuring complex physical dynamics.”
Moreover, they claimed that these efforts are for the betterment of Facebook AI; which is, of course, for the betterment and the improvement of the user experience of Facebook users.
Read: Here are the Top Ways to Keep Your Facebook Account Safe From Hackers
These efforts are part of Facebook AI’s long-term goal of building intelligent systems that can intuitively think, plan and reason about the real world … where even routine conditions are highly complex and unpredictable.”
The first audio-visual simulator which Facebook publicly announced will also add realistic simulations of any sound source. This means that the Facebook AI will have no difficulties to move through an unmapped environment to find a sound-emitting object such as a ringing phone as the social media giant likewise revealed an indoor mapping too.
AI systems could utilize and optimize to better understand and recall details regarding indoor spaces; including how many chairs are in a dining room or whether a cup is on a counter.
The Facebook AI can now follow through a command of “go find the ringing phone” rather than “go to the phone that is 25 feet southwest of your current position” as the AI is capable of finding the position of the phone on its own using sensors.
Read: Elevated Play Philippines is Hiring | Work at Home, Play at Home
With this approach, Facebook said that they were able to achieve faster, better, and more developed training and higher accuracy in navigation.
Facebook Hate Speech
Other than the actual actions and commands that people can use for it; Facebook is also on the move to trace and identify hate speech on their platform. The Facebook AI will be trained and will be the overall AI that we have been waiting for.
They are gambling through its AI system to detect hate speech on its platform. Once effective; the Facebook AI can be found in numerous servers throughout the company’s data centers.
Complex neural networks have been trained to detect and determine toxic user content whenever a new post appears on Facebook. There are those which look for hate speech; while others look for misinformation, nudity, and so on.
Bonus Read: Instagram Reels – Keeping You Off of TikTok
A wide majority of the content is forwarded to human moderators for further action. However, some of it can be analyzed and then removed by AI alone. Furthermore, Facebook also made rapid progress in the second quarter of 2020; with 104.6 million pieces of negative content (excluding spam) taken down.
Facebook’s CTO, Mike Schroepfer; said that they are hopeful that the Facebook AI will be as good or even better than a human.
We’re getting to the point where most of our systems are probably close to, as good [as], or possibly better than an untrained person in that domain. My goal is to get us to as good as experts . . . We’ll get there as fast as we can.”
Introducing the SemanticMapnet
Apart from the primary Facebook AI; the company also introduced and released the SemanticMapnet. What is SemanticMapnet? It is a tool that would help create maps of indoor spaces that are at pixel level. This tool is used to help and assist robots understand and identify them easily.
Read: How to Activate Facebook Desktop Dark Mode Feature
Some of the most concrete and the most common things that the SemanticMapnet can help solve would be identification of a particular room. For instance, you will be able to solve and answer questions about your particular space such as: “How many chairs are in the dining area?” “How many pieces of furniture are in the lobby of the office?”
The Facebook AI is a work in progress. Once it’s done, we would be able to use it all to our advantage. What do you think of the development being made by Facebook? Do you think that this will be what can actually help us?