Connect with us


The marvel of computer vision technology in today’s camera-based AR systems – TechCrunch



15 mins ago


British science fiction writer, Sir Arther C. Clark, once said, “Any sufficiently advanced technology is indistinguishable from magic.”

Augmented reality has the potential to instill awe and wonder in us just as magic would. For the very first time in the history of computing, we now have the ability to blur the line between the physical world and the virtual world. AR promises to bring forth the dawn of a new creative economy, where digital media can be brought to life and given the ability to interact with the real world.

AR experiences can seem magical but what exactly is happening behind the curtain? To answer this, we must look at the three basic foundations of a camera-based AR system like our smartphone.

  1. How do computers know where it is in the world? (Localization + Mapping)
  2. How do computers understand what the world looks like? (Geometry)
  3. How do computers understand the world as we do? (Semantics)

Part 1: How do computers know where it is in the world? (Localization)

Mars Rover Curiosity taking a selfie on Mars. Source:

When NASA scientists put the rover onto Mars, they needed a way for the robot to navigate itself on a different planet without the use of a global positioning system (GPS). They came up with a technique called Visual Inertial Odometry (VIO) to track the rover’s movement over time without GPS. This is the same technique that our smartphones use to track their spatial position and orientation.

A VIO system is made out of two parts.

Source link

Check Also

China’s new rules on video games, introduced last month, are having an effect on the …

Source link

Continue Reading
Click to comment

Leave a Reply

Your email address will not be published.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


“Warp speed,” “Prime Directive” predate Star Trek, per new reference tool



Enlarge / The term “warp drive” actually predates its first use in the long-running Star Trek franchise by 14 years.

There’s no denying the profound influence that the Star Trek franchise has had on our shared popular culture. But it turns out that some of the best-known terms associated with the series—transporter, warp speed, and the famous Prime Directive—actually predate Star Trek: The Original Series by a decade or more. According to Jesse Sheidlower, a lexicographer and editor of the newly launched online Historical Dictionary of Science Fiction (HDSF), the first mention of those terms appeared in 1956, 1952, and 1940, respectively.

The origins of this new online resource date back to 2001, when Sheidlower was working for the venerable Oxford English Dictionary (OED). “OED has always been a crowdsourced entity,” Sheidlower told Ars. “In fact, it was probably the first crowdsourced thing.” Back in the late 19th century, OED editors typically placed notices in newspapers and magazines asking people to read various materials and contribute to their coverage of the English language.

While at OED, Sheidlower noted that science fiction was an area that was not very well served by scholarship, partly because science fiction hasn’t had much serious literary cache historically. That meant that the most significant (and rare) pulp magazines weren’t available in the usual archives, like the Library of Congress or the New York Public Library. So he set up a Science Fiction Citations Project (SFCP) and called on the science fiction community (fans and writers alike) to submit examples of the specialized terminology they found, all curated by moderators.

There was particular interest on “antedatings,” the earliest known examples of given words, which are of great interest to scholars. For instance, the OED had an entry for mutant—”in the sense of a person with unusual abilities or appearance that has arisen by a genetic mutation”—dating to 1954, but Sheidlower thought it was likely coined much earlier. He was right: the first appearance of the term was in 1934. The site remained active for many years, and one of the moderators, Jeff Prucher, even published a Hugo Award-winning book, Brave New Words, in 2007.

Over time, however, all that activity waned, and the site became effectively static. Sheidlower left OED in 2013 and no longer had access to the SFCP. Last year, he asked OED for permission to revive the project, including a major design overhaul. The pandemic meant he had sufficient time to undertake such a massive overhaul, and the fact that many rare science fiction sources had by now been digitized—including the original pulps—made it easier for him to do his own extensive research.

A sample entry for "chrononaut" in the Historical Dictionary of Science Fiction.
Enlarge / A sample entry for “chrononaut” in the Historical Dictionary of Science Fiction.

Jesse Sheidlower/HDSF

Like its predecessor, the HDSF helps improve and expand our knowledge of antedatings. Thus far, according to Sheidlower, the HDSF has found more than 400 antedatings. For instance, thought-controlled was thought to date back to 1977, but it has now been traced back to 1934. Deep space dates back to 1921 (instead of 1937), ray gun first appeared in 1923, deflector was first mentioned in 1931, and the notion of a mad scientist can be traced back to 1893. And the scientific terms biotechnician and graviton were first coined in science fiction, in 1940 and 1929, respectively.

The new HDSF also included some useful added features; it’s not just a list of words. Sheidlower went to great lengths to include links whenever possible to online source material—all added manually. For example, click on mutant, and you’ll find yourself on the entry page with a chronology of its usage starting with earliest mention to current usage. Click on the 1934 mention, and it will pull up an image of the actual page where the word first appeared.

“It’s turned from a site that was basically a notebook where people could write down the research they did into something that’s more broadly useful, making it both more fun to explore and more useful to those using it for serious scholarly efforts,” said Sheidlower.

There are currently no plans to adapt the HDSF into a print book, a la Brave New Words. But Sheidlower is hoping to continue to expand the resource, particular to include more 21st-century science fiction terminology. It already includes terminology from fandom communities (-con, faan, sercon), criticism, and particularly influential science fiction films and television shows (lightsaber, redshirt, TARDIS).

And while the HDSF has no official affiliation with the OED, apart from the association of its origins, at least one OED editor approves of the project. Executive Editor Peter Gilliver described it to The New York Times as “quite impressive, and very stylishly presented,” adding, “Jesse doesn’t like to leave any stone unturned. He’s a very dogged researcher.”

Continue Reading


The complete moron’s guide to GameStop’s stock roller coaster



Enlarge / Pop goes the bubble…

Last week, an epic short squeeze had driven GameStop stock up to $40 a share, a roughly 1,500 percent increase from its low point nine months ago. Little did anyone know at the time that this would only be the beginning of the story.

As I write this, GameStop’s stock price is hovering around $350, up another 775 percent or so since I wrote about this situation eight days ago. By the time you read this, that number may be horribly outdated, as the stock continues to bounce up and down with extreme volatility hour by hour (it dipped down as low as $61 and peaked as high as $159 on Friday).

The current stock price now gives the company a market cap of about $26 billion.

On the surface, that means the market currently thinks GameStop is worth more than twice as much now (during a potentially existential threat to brick and mortar game sales) as it was during the height of the Wii boom in late 2007, when console game downloads were barely a thing.

It’s an understatement to say that nothing has changed about GameStop’s fundamental business to justify such a quick and dramatic rise in valuation. But getting at what is causing the nearly vertical launch of GameStop’s stock value is a little complicated.

A short lesson on shorts

To understand what’s happening to GameStop stock, first you have to understand short selling, where investors make a bet that a stock will go down instead of up. To do this, they borrow a share of the stock (for a fee), immediately sell it to pocket the current value, and agree to buy another share later to “cover” their short position.

But shorting stocks comes with huge risks if the stock price goes up. When your short position eventually comes due, you’re forced to buy the stock at whatever price the market currently sets, and there’s theoretically no limit to how high it could go. If you invest $1,000 in buying a stock, all you can lose is $1,000. If you borrow $1,000 worth of stock to short it, you could lose a lot more than that when you’re forced to buy much more expensive stock.

Investors as a whole were so sure that GameStop stock was going to go down that they wanted to borrow every single available share (and then some) to make money on the coming collapse.

When investors do lose money on a short position in this way, they often reborrow more short options at the new price to hedge their bet (if they expected the stock was overvalued before, it’s probably even more overvalued now, right?). That “short squeeze” process can theoretically keep going until the stock price actually starts going down (perhaps because traditional “long” investors want to pocket their profits) or the short sellers run out of capital to keep borrowing.

GameStop has been one of the most heavily shorted stocks on the market for a while now. Since the middle of 2019, the “short interest” (i.e., the number of shares investors have expressed interest in borrowing for a short position) has heavily surpassed the stock’s “free float” (i.e., the number of actively traded shares available). In other words, investors as a whole were so sure that GameStop stock was going to go down that they wanted to borrow every single available share (and then some) to make money on the coming collapse. And keep in mind, this phenomenon started when GameStop stock was already trading at historic lows of $5 a share or less.

GameStop faces plenty of headwinds to its core business of selling disc-based games in brick-and-mortar stores (as we’ve written about extensively). But that extreme level of short interest in an already heavily depressed stock was probably overly pessimistic about the company’s near-term prospects.

“They don’t have net debt, so they’re not going bankrupt or anything,” Wedbush Morgan analyst Michael Pachter told Ars last week. “And with the new console launch, they’re probably going to sell a lot of consoles and be fine.”

As it turns out, a group of retail investors noticed this excess of pessimism and was poised to exploit it.

Enter the Redditor

The WallStreetBets subreddit (WSB) describes itself as “like 4chan found a Bloomberg terminal,” and that’s not a bad description. WSB is a generally disorganized mess of posters throwing up memes and slang that can be hard to parse for an outsider. At the center of it, though, are members who analyze the market for opportunities so they can try to rally the sub’s millions of subscribers (2.9 million as of this writing) to a potential value play.

A few posters on WSB began to notice GameStop as a potentially undervalued and overshorted stock back in 2019, without too much wider action. But the group’s attention began to crystallize around April 2020, when WSB member Senior_Hedgehog laid out what turned out to be a prescient case for “the biggest short squeeze of your entire life.”

GameStop stock remained relatively flat in the months following that post. That changed in August, when founder Ryan Cohen bought a 9 percent stake in the company. Since then, Cohen has increased his stake in the company, earned control of three board seats, and promised to try to refocus the retailer into a more purely digital business.

“GameStop’s challenges stem from internal intransigence and an unwillingness to rapidly embrace the digital economy,” Cohen said in a November SEC filing. “If GameStop takes practical steps to cut its excessive real estate costs and hire the right talent, it will have the resources to begin building a powerful e-commerce platform that provides competitive pricing, broad gaming selection, fast shipping, and a truly high-touch experience that excites and delights customers.”

The YOLO-riffic logo for r/WallStreetBets.
Enlarge / The YOLO-riffic logo for r/WallStreetBets.

Whether or not that transformation will be possible, the interest from the usually risk-averse Cohen was enough to get some investors to take another look at buying GameStop. The price started creeping up from $5 in the middle of August to nearly $20 by the end of the year. That put some pressure on all those short positions, but not enough to stop them from reborrowing and keeping short interest high.

On Reddit, meanwhile, the hype around the turnaround only got louder and louder as the stock price increased. Some WSB members (and others) posted lengthy analyses suggesting a “fair value” of $100 or more per share for GameStop based on expected earnings (analysts in aggregate suggest a much more modest $13.44 price target based on the fundamentals). Others take a more meme-centric, “for the lols” route, reminding their compatriots that “stocks only go up” and urging “diamond hands” (i.e., hands that never stop holding) and bitcoin-style HODL strategies as GameStop stock goes on a “rocket to the Moon.” Some even encouraged GameStop stock owners to tell their brokers not to allow those shares to be borrowed for further short options, squeezing short borrowers even further (it’s not clear how much effect this part of the effort has had, in aggregate).

Through it all, the WSB-ers have been able to coalesce around a vision of themselves as the “little guy” using cheap retail trading tools to fight against a common perceived enemy: massive hedge funds that were heavily shorting the stock. Citron Research founder Andrew Left, who’s been publicly arguing for a $20 price target for GameStop, said last week he was targeted by a wave of harassment and attempted hacking ahead of a planned video arguing for the short position.

More recently, Melvin Capital Management has seen its value fall 30 percent this month, thanks in no small part to a heavy short position in GameStop. The hedge fund was forced to take a $2.75 billion cash infusion from Citadel to stay solvent in recent days. CNBC reports that Melvin finally closed out its short position Tuesday afternoon, i.e., taking a huge loss rather than redoubling on the short borrowing. But many WSB posters are publicly doubting that report and are confident they can literally bankrupt a massive hedge fund if the GameStop stock price goes high enough.

Continue Reading


Plex adds game-streaming as paid add-on, completely botches the landing



The streaming-hub app Plex has long endeared itself to a certain kind of media consumer—one who’d prefer to rip and stream their own purchased content from a home computer instead of relying on subscriptions. But while Plex has experimented with new features and media options over the past decade-plus, this week sees the service take its first-ever stab at interactive media. As in, games.

Plex Arcade is now live as a paid add-on service, and it promises to let you stream your existing, classic gaming library from a home PC to a variety of Plex-compatible devices. That means you can put your favorite classic game ROMs next to your music, TV, and film libraries, and you can beam those to compatible devices, either at home or afar. For some users, this is a dream scenario: centralize your classic games on one device, then access them everywhere, instead of having to manually set up each phone and set-top box with emulators and ROMs.

But Plex Arcade’s first day is a bumpy one, and even if the service patches things up, it’s a clear reminder why streaming classic games from a hub computer isn’t nearly as handy as streaming other media.

It appears you are looking for the sub-FAQ

We should begin with the service’s biggest red flag: a $4.99 monthly fee, or $2.99/mo if you already pay for Plex Pass (which comes in monthly, annual, and “lifetime” payment tiers). While we understand charging for the add-on service if you’re in Plex’s free tier, the added charge for existing Plex Pass subscribers throws us for a loop. One way Plex explains this charge is that the service comes with 27 Atari-published games for arcades and consoles, with the usual Atari hits (Asteroids, Missile Command, Centipede) along with a few end-of-an-era 7800 curios (Basketbrawl, Fatal Run, Motor Psycho).

But Plex Arcade’s monthly fee includes zero hints about a growing library of licensed arcade classics (a la Nintendo Switch Online’s slowly growing library). The cost for those users, at that point, becomes a hope that Plex Arcade delivers a pristine, drop-and-play experience, much like you can expect after ripping a TV series’ DVDs and then accessing them through an auto-sorted Plex interface on a range of Plex-supported platforms (Amazon Fire TV, Apple TV, Android TV, iOS, Android). But Plex Arcade doesn’t quite work that way.

The service’s sales pitch is as follows: “Browse old-school favorites from the Atari library or start adding your own game ROMs, then let the gaming begin!” The latter part sounds appealing: our own ROMs (which, in this day and age, are easier to self-rip or purchase as part of licensed ROM collections). Why wait for a rights handler to get my favorite SNES or Game Boy classic titles dumped to a modern platform when Plex can help me serve my own ROMs from my computer to other devices? This sales pitch hints to an instant process, and you’d be fooled into thinking it’s that simple when you point your Plex Media Server app to a directory full of your own game ROMs (after installing Parsec as a prerequisite). When I did this, I waited for the scanning process to conclude, and then I found my Plex mobile app had gotten some nifty new tabs: “Arcade,” “Game Boy,” “Nintendo Entertainment System,” “Sega Genesis,” “Super Nintendo Entertainment System,” and “Nintendo 64.” (ROMs from my GameCube, Wii, Xbox, and PlayStation 1 folders did not get picked up.)

Excitedly, I tried booting games from each of those platforms on my smartphone, only to get an endlessly looping “loading” animation on a black screen. Confused, I went into the Windows interface to see if I could boot the games there, at which point I was met with an error message: “The core mapping file is missing a core for [console name].” Good thing I did this on my Windows rig, since this at least pointed me to the problem; the lack of verbose warnings on Plex’s smartphone app seems like a glaring omission at this early stage in the service’s life.

I looked through the service’s FAQ and felt confused, so I reached out to Plex representatives, who pointed me to a sub-FAQ about emulators. As it turns out, while Plex Arcade is smart enough to recognize ROMs and immediately silo them into respective “platform” tabs, it otherwise forces users to procure their own emulation cores—specifically ones built for the RetroArch interface. Plex goes so far as to suggest that users download RetroArch outright and use its automatic core-update system, then drag-and-drop those core DLL files into Plex’s interface (and this includes exposing the “AppData” folder that Windows hides by default, which Plex doesn’t clarify for brand-new users). Then you have to manually edit an XML file and use precise metadata tags to tell Plex Arcade that you indeed downloaded those cores. (A single typo could nuke the whole file.)

Like, why can’t you do all of that, Plex?

Issues with customization, latency

In better news, Plex Arcade does a decent job grabbing metadata about the ROMs you load into the interface, complete with splash images and sortable information (release year, publisher, genre), though the gathered information is hit-or-miss—and apparently sourced from, which has its fair share of typos depending on which game you’re looking up. And the app plays nicely with the usual gamut of Bluetooth gamepads, along with direct-connection controller options like the Razer Kishi.

Once you’re inside a game, the issues continue to mount. Most severely, Plex Arcade offers zero useful emulation options. Want to remap controls (especially with N64 games, which do not map to Xinput)? Change rendering resolution? Choose a screen ratio? Add or modify visual filters? Create midgame “save states,” which you can then transfer from one Plex device to another? The answer to all of those is “no.” (To the point about save states: Plex Arcade does seem to support “freezing” midgame and coming back later, but I’ve seen this crash inconsistently, and there’s no way to manually trigger a reliable save state.) There’s also the matter of screen resizing not being customizable depending on whether you use a connected gamepad or on-screen controls. In some games, like the included Super Breakout arcade version, the on-screen control grid obscures the gameplay if you play in vertical orientation.

Perhaps worst of all, as of press time, the on-screen control system maps its “d-pad” to RetroArch’s d-pad functionality… which means none of the included Atari games work. Whether from Atari’s consoles or arcade releases, these 27 games require joystick mapping, not d-pad mapping, within the RetroArch ecosystem. We hope you enjoy only shooting guns and propelling directly forward in Asteroids, because until we hear otherwise, steering is out of the question.

Streaming from a Plex server to a phone makes the most sense if you’re itching to play high-end games that a phone can’t necessarily emulate. But Plex Arcade is inherently limited to classic gaming’s simplest consoles—probably because newer consoles often require ripping a BIOS file (or coding your own) to work, and that crosses a legal line. Plex Arcade’s supported emulation platforms don’t need to go so far. The most “advanced” supported system, N64, already has a range of solid emulators designed for mobile SoCs.

Thus, beyond a lack of configuration options, Plex Arcade’s sales pitch is pretty laughable from a gameplay standpoint. The smartest thing to do with twitchy, classic gaming on the go is to directly install emulators and ROMs, and Plex Arcade’s supported platforms are all (for the most part) easy enough to jailbreak or sideload emulation content onto. RetroArch has long been optimized for Snapdragons and other mobile-friendly SoCs in the world.

Thanks to issues with getting RetroArch’s cores loaded, I could only test the included pack of Atari games during the service’s first, buggy day. As in, the simplest fare imaginable. Yet these games all suffered noticeably from input lag while trying to use the Plex app on an Android phone connected to my router’s 5GHz band. All of this, for an Atari game library that doesn’t exceed 1.2MB of on-disk storage. (Hilariously, Plex Arcade doesn’t bother encrypting its 27-game Atari library, so you can conceivably sign up for the service’s seven-day free trial, copy-and-paste the 1.2MB of files elsewhere, then cancel.)

Not a good sign for a paid service

Plex Arcade could conceivably smooth out the aforementioned issues in future updates, particularly in exposing each RetroArch core’s range of options. And anybody who’s already a Plex user can understand the appeal: beam classic ROMs to supported Plex devices, and trade some latency for the convenience.

But the fact of the matter is, Plex Arcade launched with a bonkers requirement that users pick through hidden directories, manually edit an XML file, and procure their own RetroArch cores, should they wish to go beyond the service’s 27 built-in games. And those users get to pay a monthly fee for that privilege. That public, paid launch plan doesn’t make me confident that Plex has a simpler plug-and-play plan in store, at least not in the near future.

Hence, until Plex convinces us otherwise, we’ll continue to direct interested retro gamers toward RetroArch installs on their favorite mobile devices or set-top boxes. And even if you’re interested, perhaps wait to use the service’s one-time, seven-day free trial until Plex Arcade gets more of its feature set sorted out.

Continue Reading