The out-there AI ideas designed to keep the US ahead of China
President Trump signed an executive order last month to launch the US government’s AI strategy, called the American AI Initiative. Kratsios, who is also the government’s deputy chief technology officer, has been the driving force behind White House strategy on AI. The American AI Initiative calls for more funding and will make data and computing resources available to AI researchers.
But many of the algorithms now considered AI were developed many years ago, and they are fundamentally limited. “We are harvesting the intellectual fruit that was planted decades ago,” says John Everett, deputy director of DARPA’s Information Innovation Office. “That’s why we’re looking at far forward challenges—challenges that might not come to fruition for a decade.”
Through its AI Next program, DARPA has launched nine major research projects meant to tackle those limitations. They include a major effort to teach AI programs common sense, a weakness that often causes today’s systems to fail. Giving AI a broader understanding of the world—something that humans take for granted—could eventually make personal assistants more helpful and easier to chat with, and it could help robots navigate unfamiliar environments.
Another DARPA project will seek to develop AI programs that learn using less data. Training data is the lifeblood of machine learning, and algorithms that can ingest more of it can leap ahead of the competition. An innovation in this area could knock out a key advantage of tech companies operating in China, for example, which thrive on their access to an abundance of data. Other projects being funded focus on designing more efficient AI chips; exploring ways to explain the decision-making of opaque machine-learning tools; and making AI programs more secure.
See the full story here: https://www.technologyreview.com/s/613089/the-out-there-ai-ideas-designed-to-keep-the-us-ahead-of-china/
Q&A: The Future Group’s Pixotope aims to shake up augmented and virtual production
How do you see it fitting into the larger AR/VR ecosystem?
Whilst the broadcast market is currently our most important market, it’s clear that cross-reality content, particularly augmented reality (AR), will permeate all market segments in the long run.
We, therefore, needed to create a solution that would address the requirements of other market segments as well, including corporate, industrial, medical, educational, digital marketing, and events.
Consequently, Pixotope has been specifically designed to be simple to use, really fast to design with, it connects to third-party products easily for added functionality, and it outputs graphics content of the highest possible quality today …even at 4K, and later in 2019, HDR too.
The Future Group will showcase Pixotope during the upcoming 2019 NAB Show in stand SU12702.
Brain Computer Interface for Neuro-rehabilitation With Deep Learning Classification and Virtual Reality Feedback
Though Motor Imagery (MI) stroke rehabilitation effectively promotes neural reorganization, current therapeutic methods are immeasurable and their repetitiveness can be demotivating. In this work, a real-time electroencephalogram (EEG) based MI-BCI (Brain Computer Interface) system with a virtual reality (VR) game as a motivational feedback has been developed for stroke rehabilitation. If the subject successfully hits one of the targets, it explodes and thus providing feedback on a successfully imagined and virtually executed movement of hands or feet. Novel classification algorithms with deep learning (DL) and convolutional neural network (CNN) architecture with a unique trial onset detection technique was used. Our classifiers performed better than the previous architectures on datasets from PhysioNet offline database. It provided fine classification in the real-time game setting using a 0.5 second 16 channel input for the CNN architectures. Ten participants reported the training to be interesting, fun and immersive. "It is a bit weird, because it feels like it would be my hands", was one of the comments from a test person. The VR system induced a slight discomfort and a moderate effort for MI activations was reported. We conclude that MI-BCI-VR systems with classifiers based on DL for real-time game applications should be considered for motivating MI stroke rehabilitation.
See the full post here: https://dl.acm.org/citation.cfm?id=3311864
Apple’s smart glasses can change the game in a niche augmented & virtual reality market
We don't need more technology, we need someone to find a true use for Virtual and Augmented Reality. Apple is just the firm to do it —since it's done the same thing twice before.
See the full story here: https://appleinsider.com/articles/19/03/09/apples-smart-glasses-can-change-the-game-in-a-niche-augmented-virtual-reality-market
In a note to investors on Friday, TF International Securities analyst Ming-Chi Kuo said that Apple will start producing its AR headset as soon as the fourth quarter of 2019. Kuo cautioned, however, that mass production could slip to the second quarter of 2020, if development doesn’t go as planned, according to Apple-tracking site 9to5Mac, which obtained a copy of the note.
According to Kuo, Apple’s AR glasses will rely on the iPhone to power the features and will not act independently. Android users, then, will either need to buy an iPhone to use Apple’s device or find an alternative.
http://fortune.com/2019/03/08/apple-augmented-reality-headset-2/
High Fidelity’s Multi-Con VR Event Will Feature $10K Avatar Contest
High Fidelity this week announced the launch of Multi-Con VR, a comic con-style online VR event featuring contests, cosplay, fan art, celebrity panelists, and a slew of other exciting happenings.
The highlight of this event however, will be yet another Avatar Cosplay Contest. This time around, High Fidelity is taking their cut-throat competition even further with a grand prize of $10,000 USD for first place. Second place isn’t anything to scoff at either, with the runner-up taking home a respectable $5,000. Third place will earn themselves a solid $1,000, the same amount as FUTVRE LANDS first place winnings. There’s even a generous participation award of $300for every competing avatar.
A prizes will be paid in High Fidelity Coin, which can be cashed out and exchanged for standard US currency.
Contestants looking to take home gold will first need to impress the competitions panel of judges, composed of the build team from the Emmy-nominated television show, MythBusters.
Multi-Con will take place across three separate dates, each of which featuring their own unique topic. March 16th will revolve around anime/animation and the Avatar Cosplay Contest. April 27th will revolve around fantasy and sci-fi, and June 8th will be reserved specifically for comic book lovers.
Multi-Con is a free event, however you will need to apply for the avatar contest here before being considered for competition.
Valve Psychologist to Explore Brain-Computer Interface Research at GDC
At GDC 2019 later this month, Valve’s Principal Experimental Psychologist, Mike Ambinder will present the latest research pertaining to brain-computer interfaces—using signals from the brain as computer input. Ambinder says that BCI is still “speculative technology,” but could play an important role in the way players interact with the games of the future.
As time moves forward, the means by which users interact with computers have becoming increasingly natural.
While we’re far, far away from anything like direct brain I/O, there has been some headway made in recent years at least on the input side—’brain reading’, if you will. And while early, there’s exciting potential for the technology to transform the way we interact with computers, and how computers interact (and react) to us.
At GDC 2019 later this month in San Francisco, Valve’s Principal Experimental Psychologist, Mike Ambinder, will present an overview of recent BCI research with an eye toward its applicability to gaming. The session, titled Brain-Computer Interfaces: One Possible Future for How We Play, will take place on Friday, March 22nd.
Don’t break up big tech — regulate data access, says EU antitrust chief
“Because I still find that it’s quite tricky to understand what it is that you accept when you accept your terms and conditions. And I think it would be great if we as citizens could really say ‘oh this is what I am signing up to and I’m perfectly happy with that’.”
Breaking up tech giants should be a measure of last resort, the European Union’s competition commissioner, Margrethe Vestager, has suggested.
“To break up a company, to break up private property would be very far reaching and you would need to have a very strong case that it would produce better results for consumers in the marketplace than what you could do with more mainstream tools,” she warned this weekend, speaking in a SXSW interview with Recode’s Kara Swisher. “We’re dealing with private property. Businesses that are built and invested in and become successful because of their innovation.”
Vestager has built a reputation for being feared by tech giants, thanks to a number of major (and often expensive) interventions since she took up the Commission antitrust brief in 2014, with still one big outstanding investigation hanging over Google.
But while opposition politicians in many Western markets — including high profile would-be U.S. presidential candidates — are now competing on sounding tough on tech, the European commissioner advocates taking a scalpel to data streams rather than wielding a break-up hammer to smash market-skewing tech giants.
The commissioner has spoken up before about regulating access to data as a more interesting option for controlling digital giants vs breaking them up.
And some European regulators appear to be moving in that direction already. Such the German Federal Cartel Office (FCO) which last month announced a decision against Facebook which aims to limit how it can use data from its own services. The FCO’s move has been couched as akin to an internal break up of the company, at the data level, without the tech giant having to be forced to separate and sell off business units like Instagram and WhatsApp.
It’s perhaps not surprising, therefore, that Facebook founder Mark Zuckerberg announced a massive plan to merge all three services at the technical level just last week — billing the switch to encrypted content but merged metadata as a ‘pro-privacy’ move, while clearly also intending to restructure his empire in a way that works against regulatory interventions that separate and control internal data flows at the product level.
Asked for her worst case scenario for tech 10 years hence, she said it would be to have “all of the technology but none of the societal positive oversight and direction”.
On the flip side, the best case would be for legislators to be “willing to take sufficient steps in taxation and in regulating access to data and fairness in the marketplace”.
“We would also need to see technology develop to have new players,” she emphasized. “Because we still need to see what will happen with quantum computing, what will happen with blockchain, what other uses are there for all if that new technology. Because I still think that it holds a lot of promise. But only if our democracy will give it direction. Then you will have a positive outcome.”
See the full story here: https://techcrunch.com/2019/03/11/dont-break-up-big-tech-regulate-data-access-says-eu-antitrust-chief/
Niantic’s Harry Potter: Wizards Unite is a sorcerous smorgasbord for the Pokémon GO generation
Niantic’s follow-up to the absurdly popular Pokémon GO, the long-awaited Harry Potter: Wizards Unite, has one major drawback: unlike its predecessor, you can’t explain it in a single sentence. There’s so much to do in this game that it may repel some casual players — but while its depths of systems and collectibles may be nigh endless, don’t worry: you still basically just walk around doing wizard and witch stuff.
...it’s coming out sometime in 2019 (I’d guess before Summer but they wouldn’t say), in 17 languages (listed at bottom; actual countries where it’ll be offered unknown), there’s no wand accessory yet (I asked and they all looked nervous), minimum specs are reasonable and AR is optional, and it’s free but there are in-app purchases.
In HP:WU you:
- Walk around a wizarding-themed version of the real world looking for locations at which to resupply and “foundables” to encounter
- Dispel, battle, or otherwise deal with the “confoundable” associated with these
- Earn reward items from encounters and for entering foundables in your registry
- Use reward items to level up in various professions, brew potions, and battle alongside others at “fortresses”
- Find rare foundables that advance the overall plot of why this is all happening anyway
See the full, lengthy article here: https://techcrunch.com/2019/03/11/niantics-harry-potter-wizards-unite-is-a-sorcerous-smorgasbord-for-the-pokemon-go-generation/
Real-Time AR Self-Expression with Machine Learning
One of the key challenges in making these AR features possible is proper anchoring of the virtual content to the real world; a process that requires a unique set of perceptive technologies able to track the highly dynamic surface geometry across every smile, frown or smirk.
To make all this possible, we employ machine learning (ML) to infer approximate 3D surface geometry to enable visual effects, requiring only a single camera input without the need for a dedicated depth sensor. This approach provides the use of AR effects at realtime speeds, using TensorFlow Lite for mobile CPU inference or its new mobile GPU functionality where available.
An ML Pipeline for Selfie AR
Our ML pipeline consists of two real-time deep neural network models that work together: A detector that operates on the full image and computes face locations, and a generic 3D mesh model that operates on those locations and predicts the approximate surface geometry via regression.
Once the location of interest is cropped, the mesh network is only applied to a single frame at a time, using a windowed smoothing in order to reduce noise when the face is static while avoiding lagging during significant movement.
The 3D mesh network receives as input a cropped video frame. It doesn't rely on additional depth input, so it can also be applied to pre-recorded videos. The model outputs the positions of the 3D points, as well as the probability of a face being present and reasonably aligned in the input.
Hardware-tailored Inference
We use TensorFlow Lite for on-device neural network inference.
The end result of these efforts empowers a user experience with convincing, realistic selfie AR effects in YouTube, ARCore, and other clients by:
- Simulating light reflections via environmental mapping for realistic rendering of glasses
- Natural lighting by casting virtual object shadows onto the face mesh
- Modelling face occlusions to hide virtual object parts behind a face, e.g. virtual glasses, as shown below.
In addition, we achieve highly realistic makeup effects by:
- Modelling Specular reflections applied on lips and
- Face painting by using luminance-aware material
See the full story here: https://ai.googleblog.com/2019/03/real-time-ar-self-expression-with.html
Nintendo Unveils Virtual Reality Cardboard Headset for Switch
The Nintendo Labo VR Kit lets gamers build a headset mostly out of cardboard so that they can slide in the device to create what the company calls “basic VR technology.” It’s reminiscent of Google’s Cardboard, which turns smartphones into VR goggles and was introduced almost five years ago. Priced at $80, Nintendo’s VR kit includes an alien shooting title and an ocean swimming simulation. It goes on sale April 12.
“We wanted to design an experience that encourages both virtual and real-world interactions among players,” Doug Bowser, Nintendo of America’s chief, said in a statementThursday.
Nintendo’s VR headset doesn’t require a strap-on to stay affixed. Instead, users hold it up to their eyes with one hand and control games with the other. That suggests a focus on more casual, shorter experiences and a device that’s meant to be easily shared among a group. Spending on VR and augmented reality will grow 69 percent this year to $20.4 billion, with hardware accounting for half of that, according to IDC.
The new product is Nintendo’s first move into VR in more than two decades, after the 1995 Virtual Boy flopped with consumers.
See the full story here: https://www.bloomberg.com/news/articles/2019-03-07/nintendo-unveils-virtual-reality-cardboard-headset-for-switch
Pages
- About Philip Lelyveld
- Mark and Addie Lelyveld Biographies
- Presentations and articles
- Trustworthy AI – A Market-Driven approach
- Tufts Alumni Bio
