• Sat. Sep 30th, 2023



The neural network forced portraits to speak, Samsung learned to copy the brain and other news from the world of AI


Nov 22, 2022

The neural network forced portraits to speak, Samsung learned to copy the brain and other news from the world of AI

We strive to inform readers not only about events in the Bitcoin industry, but also to talk about what is happening in related technological spheres-cybersecurity, and now in the world of artificial intelligence (AI).

Forklog has collected the most important news from the world of AI over the past week.

  • Amazon introduced new devices and services with artificial intelligence.
  • Yandex added a real -time recognition function using a smartphone camera.
  • Samsung offered to “copy and insert” the brain into neuromorphic chips.
  • Cruise and Waymo allowed to operate robotax in San Francisco.
  • AI denied the authenticity of the famous picture of Rubens “Samson and Dalila”.
  • Texas police sued Tesla because of an accident with a drone.
  • The neural network added Beethoven’s symphony.

Amazon introduced new devices with artificial intelligence

On September 28, 2021, Amazon introduced a number of products with artificial intelligence.

Among the new products was the Astro home robot based on the virtual assistant Alexa. He knows how to recognize faces, respond to voice teams, transfer things and observe what is happening around.

The smart display ECHO Show 15 is also presented with support for detecting certain sounds and Visual ID for user identification.

Amazon has updated some ALEXA confidentiality functions that allow AI assistant to process data locally on devices without transmission to the cloud.

The company also showed the home drone Always Home Cam with the detection of traffic to ensure safety and virtual guard Virtual Security Guard to control street cameras and monitor activity monitoring.

The release date of new products will be reported later.

Yandex added the function of recognizing objects to a mobile application

Yandex Updated the Smart Camera in the official application.

When recognizing simple food products, the system indicates the amount of fats, proteins, carbohydrates, microelements, calories, as well as the content of sugar and GMOs.

Working function “smart camera”. Data: “Yandex”.

The company added identification of cars with information about their brand, model, power, speed and fuel consumption.

In addition, the camera was taught to predict the procedure for solving linear and square equations, as well as find training videos with the analysis of such tasks.

Google will add to Android 12 brutal control of a camera based

Google has added to Android 12 telephone management functions of facial expressions and handwritten text recognition. They are intended for people with disabilities.

Camera Switches allows you to open various commands with the help of facial expressions, for example, opening the mouth or lifting eyebrows.

Project Activate reads facial expressions for playing music or sending a message.

Lookout recognizes handwritten and printed text using a smartphone camera. The function works for languages ​​using Latin. The developers also added the identification of euro and Indian rupees in foreign currency mode.

Samsung introduced the technology of brain copying into neurochip

Samsung offered to “copy and insert” the brain into neuromorphic chips.

The essence of the concept is to duplicate neural connections using the nanoelectroda array and their transfer to the module of a three -dimensional network of solid -state memory.

The authors propose to use the approach to create a microcircuit that allows low energy consumption, easy learning, adaptation to the environment, autonomy and ability to cognition.

Russian developers have created an application for transcribing media files

TAD Development has launched a service with artificial intelligence “AI Secretary” for transcribing audio and video recordings.

The system recognizes speech on media files and transfers it to the text, observing the rules of spelling and punctuation. Its accuracy of work is 80%.

The application is free and available on the company’s website, as well as in the App Store and Google Play.

Cruise and Waymo allowed to operate robotax in San Francisco

Manufacturers of drone Cruise (General Motors) and Waymo (Alphabet) received permission for commercial operation of robotaxi in San Francisco.

The California Department of vehicles allowed Cruise to make trips without a insured driver while driving, while Waymo should always be present.

Robotax Cruise can deliver passengers in certain parts of the city from 22:00 to 6:00 at a speed of not more than 48 km/h/h. They were also allowed to go to “light rain and fog”. Waymo cars were allowed to travel on public roads at a maximum speed of 104 km/h.

Police Texas filed a lawsuit against Tesla because of an accident with five victims

In Texas, law enforcement officers sued Tesla after Model X with the autopilot included in a group of police officers.

On February 27, 2021, an electric car at a speed of 70 km/h crashed into two parked officers’ cars and pushed them with patrolmen who conducted an inspection of another car. Five people received serious injuries.

The plaintiffs argue that the cause of the accident was the “constructive and production defects” of the car, as well as the inability of the autopilot to recognize police cars and warn about the dangers.

A lawsuit was put forward against DeepMind due to the use of patients in the UK

In the UK, they sued the DeepMind II laboratory due to the allegedly illegal use of medical data.

In 2015, the company and Royal Free London NHS Foundation Trust entered into an agreement within which the developers gained access to 1.6 million patients to develop applications for doctors.

On September 30, 2021, the Law Firm Mishcon de Reya filed a collective lawsuit on behalf of Andrew Famoll and other British citizens to DeepMind due to the acquisition of confidential information “without permission, knowledge or consent”.

Artificial intelligence denied the authenticity of the legendary picture

Scientists with the help of AI refuted the authenticity of Rubens’ canvas “Samson and Dalila”, represented in the National Gallery in London.

Swiss experts analyzed the picture and determined that “each of its fragments was more than 90% false”. They repeated the experiment, but the result has not changed.

The museum representative said that they always take note of new research and wait for the complete publication of work to evaluate evidence.

The London National Gallery acquired the canvas in 1980 for $ 3.4 million, which is now equivalent to $ 11.5 million, taking into account inflation.

AI-Startap taught portraits to move and talk

The Israeli startup D -id has developed a Speaking Portraits tool that revives a person in the photo and makes him speak.

To train the algorithm, the user needs to upload a text file, audio or short video.

The developers have created two tool options – Single Portrait and Trained Character. The first only revives the head, and the second generates the character with a certain set of actions and words.

The neural network added Beethoven’s symphony

A group of musicians and developers graduated from the tenth symphony of Beethoven using artificial intelligence.

The authors investigated the work of the German composer, analyzed them and digitized them. On these data, they trained a neural network, into which the drafts of an https://gagarin.news/news/the-cryptocurrency-market-capitalization-fell-below-a-trillion-dollars/ incomplete work later loaded. For 18 months, the team created and orchestrated two full parts of the tenth symphony lasting more than 20 minutes each.

Researchers will present the results of the work on October 9, 2021.

Also on FORKLOG:

  • In Russia, artificial intelligence will calculate “dangerous schoolchildren” for works.
  • The creator of Atari will release a NFT collection of classic arcade machines in augmented reality.
  • Scientists have developed an algorithm for determining animal viruses transmitted to a person.
  • Deepmind has developed an algorithm for predicting rain a few hours ahead.
  • EU and the USA will develop general principles of artificial intelligence.

Subscribe to FORKLOG news at Telegram: Forklog AI – all news from the world of AI!

Leave a Reply

Your email address will not be published. Required fields are marked *