Known globally for its contributions to the MPEG format — the compression technology used in MP3 audio files and MP4 films — Germany’s Fraunhofer has grown to become its interest to the subsequent frontier in media: virtual reality. After unveiling lower-priced VR headset microdisplay hardware last year, the employer is now displaying off next-era video compression software program the usage of the brand new MPEG-OMAF fashionable, the primary VR specification permitting 360-degree videos to flow over 5G networks.
Based in “considerable” part upon Fraunhofer video compression technologies, MPEG-OMAF breaks wraparound films into grids of tiles encoded at multiple resolutions. The explanatory picture above uses pink tiles to indicate areas that might be streamed at low resolution, as opposed to normally colored tiles, which can be streamed at high resolution. Unlike conventional videos, which move from servers to one consumer solution, those VR videos dynamically use high-resolution tiles in which the viewer is currently looking and occasional-resolution tiles for components that might be out of sight. As the consumer’s head role modifications, the headset or display tool requests an exclusive blend of streamed tiles optimized for the user’s modern recognition vicinity.
This trick allows the whole 360-degree video to continue streaming even as devoting most detail to whatever the consumer is viewing. It parallels the current use of foveated rendering to maximize real-time 3D pictures for VR users, ensuring that head-moving viewers can constantly see something through their peripheral vision, even though it decreases infidelity. International mobile standards business enterprise 3GPP has adopted the MPEG-OMAF, popular for 5G VR streaming, so it will likely underpin plenty of 360-degree virtual fact video streams — just like MP3 and MP4 defined previous generations of digital audio and video. Current 360-degree VR videos streamed over 4G and even Wi-Fi networks tend to be afflicted by low average resolutions, across-the-board compression artifacts, and excessive latency, all of which the new widespread and higher bandwidth networks could dispose of.
The actions provide fresh evidence of Apple’s renewed pursuits to enter the self-sufficient automobile derby, an effort it calls Project Titan. The talks are focused on the next-era lidar, a sensor that gives a 3-dimensional look at the street. The three people said that Apple is searching for lidar gadgets that could be smaller, less expensive, and greater without difficulty mass-produced than the modern generation. The iPhone maker is putting an excessive bar with needs for a “progressive layout,” one of the humans acquainted with the talks started. The people declined to name the groups Apple has approached.
The sensor effort means Apple desires to increase the entire hardware chain to guide self-sufficient automobiles and has joined automakers and investors in the race to locate winning technology. Current lidar systems, such as units from Velodyne established on Apple’s fleet of self-driving cars, use mild laser pulses to render specific pix of the environment around the vehicle. But the structures can be valued at $ 100,000 and use mechanical parts to brush the laser scanners across the street.
That makes them too bulky and at risk of failure to be used in mass-produced cars. The shortcomings have spurred $1 billion in investment at dozens of startups and mature agencies to make lidar smaller, inexpensive, and more robust. Apple’s interest in subsequent-technology lidar sensors comes as it has sharply extended its avenue checking out while bringing on key hires from Tesla and Alphabet’s Google. It remains doubtful whether Apple’s Project Titan aims to construct its car or supply a self-using automobile’s hardware and software program elements while pairing with a partner for the entire vehicle.
But what is apparent from Apple’s hobby in cheaper lidar structures is that it desires to manipulate the “notion stack” of sensors, computer systems, and software to pressure a self-reliant automobile, no matter who makes the automobile, some other man or woman acquainted with the talks started. The three people acquainted with the negotiations declined to be diagnosed because the discussions were not public. In addition to evaluating the capability of outside providers, Apple is thought to have its inner lidar sensor below improvement, two human beings said. Alphabet-owned Waymo has taken a comparable route, assembling a sensor and laptop system even as inking offers to shop for motors from Fiat Chrysler Automobiles. Apple receives “a variety of optionality with the aid of running at the perception stack,” the second character acquainted with the talks. “Bringing a passenger automobile to the market is truly, certainly hard, and there’s no cause proper now they need to jump into it.”
The designs Apple is seeking could probably be made with conventional semiconductor manufacturing techniques, all four human beings familiar with the talks started. That can lower fees from the various thousands to loads of bucks as the sensors are produced in large numbers, similar to chips in phones and other gadgets. Apple also needs sensors that can see several hundred meters (yards) down the street. The long-distance requirement shows Apple is interested in fully self-riding vehicles versus the greater restrained capabilities, including adaptive cruise management used today, people acquainted with the problem said. “They’re no longer glad about most of what they see,” the primary character familiar with the matter said. “They’re searching out modern design.” A 0.33 man or woman acquainted with the problem said Apple is searching for a “design-oriented” sensor that could be glossy and discreet sufficient to suit the general strains of a vehicle.