We strive to not only inform readers about events in the bitcoin industry, but also to talk about what is happening in the relevant technological fields – cybersecurity and artificial intelligence (AI) world.

ForkLog AI gathered the most important news from the world of AI last week.

Nearly 400 robot cars crashed last year

The US National Highway Traffic Safety Administration (NHTSA) has released a report on accidents involving robocars and vehicles with driver assistance systems.

According to the agency, from July 1, 2021 to May 15, 2022, 392 accidents were related to drones. In 108 cases, another vehicle, 11 other cyclists and pedestrians were involved. 11 out of 98 fatal accidents were seriously injured.

The 273 crashes involve Tesla robots using Autopilot or Fully Self-Driving. Five incidents resulted in death.

Honda vehicles have been associated with 90 crashes, Waymo 62, Transdev Alternative Services 34, Cruise 23, Subaru 10. Drones from other automakers such as Ford, General Motors, Volkswagen and Toyota have been involved in five or fewer crashes.

Tesla to buy video camera for robot cars from Samsung

Tesla has partnered with Samsung to supply robot car cameras.

According to Electrek, the tech giant won a $3.2 billion-$4 billion tender. Samsung will supply 5-megapixel version 4.0 cameras that provide five times the sharpness of 3.0 modules.

The source said that Tesla will integrate video devices into all robot cars in the next few years.

Adobe introduced a tool to restore old photos

Adobe has developed a neurofilter to restore old damaged photos in Photoshop.

The system is based on artificial intelligence Sensei AI. Neurofilter removes images of dirt spots, scratches, wrinkles and tears. The technology smoothes the skin of a person in the portrait, removing unwanted textures or noise, and adding tints.

The filter can work with other tools like Content-Aware Fill to restore missing parts of a photo.

The company also announced a free web version of the photo editor with basic features.

Meta will present a method for creating realistic 3D avatars using a smartphone at Siggraph 2022.

The approach allows you to digitize the head in four minutes and then create its three-dimensional model with a wide variety of emotions. To do this, the user must “scan” the face with the smartphone’s front camera and recreate 65 specific expressions.

According to scientists, the process of creating a 3D avatar on a high-performance computer will take about six hours. You can also use the cloud to create.

The researchers added that the method doesn’t work with glasses and makes long hair look bad.

Study: Facebook tracks people seeking abortion services

Facebook advertising tools track people seeking abortion services. Reveal said the system should filter sensitive health data, but collect it instead.

Researchers discovered that hundreds of birth crisis centers use the Meta Pixel analysis tool. The system allows organizations to track visits to sites to create targeted ads on Facebook.

According to the report, pregnancy crisis centers are generally not licensed medical facilities and are not affiliated with HIPAA. They are mostly run by religious organizations that encourage people to stop having abortions.

According to Reveal, Meta Pixel collects “highly” sensitive information, including scheduled meetings. Third-party anti-abortion marketing companies may also have access to the information collected by the tool, even if their website has not been visited.

ByteDance will invest in the development of virtual reality content

TikTok owner ByteDance plans to invest “a lot of money” in developing virtual reality content.

According to the protocol, most of the funds will go to recruiting staff for Pico, the company-owned VR headset maker. Most job postings are related to the content-oriented department of Pico Studios.

Microsoft Teams uses AI to improve audio in video conferencing

Microsoft used artificial intelligence to improve the quality of communication in Teams video chats.

The company said the new algorithms can eliminate overlaps associated with sound reflections, minimizing echo and reducing the chance of outages. Microsoft added that all processing is done locally on client devices.

According to The Verge, the developers created the technology using 100,000 simulated rooms and trained it for 30,000 hours of conversation. The tech giant also paid Teams users to record their voices to train the AI ​​to recognize thousands of different devices.

The feature comes with other updates including AI improvements and text display optimizations to limit bandwidth during video calls.

AI will examine images of space from the James Webb telescope

Artificial intelligence will be used to examine the first images from the James Webb Space Telescope (JWST). Images will be released on July 12, 2022.

A team of 50 researchers will use the trained Morpheus AI system to analyze pixels, find blurry droplets in space photographs, and determine if these structures are galaxies and what type. The model will also help study the evolution of dark matter and the beginning of the birth of stars.

Morpheus was trained on 7,600 images of galaxies taken by NASA’s Hubble Space Telescope. However, the researchers plan to retrain the algorithm to work with JWST data.

Researchers taught a neural network to create a video based on a text description

Researchers from Tsinghua University (China) have developed a CogVideo transformer with 9 billion parameters to convert text to video. The neural network only works with Chinese.

The scientists trained the algorithm by inheriting a pre-trained CogView2 word query image generator.

The model creates a 4-second video clip of 480×480 pixels at 8 frames per second.

New Voicemod filter will let you speak with Morgan Freeman’s voice

Voicemod uses AI to render Morgan Freeman’s voice in real time. The beta version of Voicemod AI Voices converter also includes seven new filters.

The developers brought in professional English-speaking voice actors with the necessary speech capabilities to create a training dataset. Audio engineers then processed the data in traditional ways, such as adding background music and dynamic effects.

Voicemod AI Voices are available on PC. Filters can be used on Twitch streams or Discord searches.

Last week’s top AI deals

From June 12 to June 18, 2022, AI startups exceeded $237 million. Here are the most interesting trades.

• Invoca received $83 million to develop software to analyze call center operators’ work on calls.
• Charm Therapeutics has raised $50 million to develop an AI platform for modeling complex protein interactions.
• Lightning AI has raised $40 million to expand its team of AI tools developers.
• For the design of Alice Technologies building structures, the developers of the platform raised $ 30 million.
• Able received $20 million to speed up the processing of customer data by banks and other lenders.
• Ahana, a managed services developer based on the Presto query engine, has raised $7.2 million.
• Workstream has raised $7 million to develop a platform to streamline enterprise analytics workflows.

Also on ForkLog: