speech
Latest
Meta's newest dataset will train speech recognition engines on 'clusters' of speakers
A newly developed dataset from Meta AI, however, promises to improve the performance of such automatic speech recognition (ASR) tools by clustering speech at the “utterance level.”
Severely paralyzed man communicates using brain signals sent to his vocal tract
A severely paralyzed man communicated using a system that translates signals from his brain to his vocal tract directly into words that appear on a screen.
Microsoft opens limited access to its neural text-to-speech AI
AT&T and Warner Bros. used the tech to create an interactive version of Bugs Bunny.
Google wants you to train its AI by lip syncing 'Dance Monkey' by Tones and I
Google is asking users to lip sync 'Dance Monkey' by Tones and I in order to train its AI.
Facebook removes hundreds more accounts over hate speech
Facebook has removed more accounts associated with hate groups.
Ivanka Trump will reportedly appear on a panel at CES
According to documents reviewed by CNET, Ivanka Trump is scheduled to speak at CES 2020. Supposedly, she's slated as a headline speaker in a panel with Consumer Technology Association (CTA) president Gary Shapiro.
The latest Xbox One update adds Google Assistant voice controls
Xbox One's November update is here, bringing with it a bunch of new features and tools designed, as always, to improve your gaming experience. First up, and probably most notable, is the roll-out of Google Assistant voice controls. We've known this was coming for a while, and after a brief spell in beta, it's now available to all Xbox users. If you've got a Google Assistant Home-enabled device, you'll be able to turn your console on and off, launch games and apps and control videos using your voice, and the usual "Hey, Google..." command.
Mark Zuckerberg defends free speech on Facebook
Facebook CEO Mark Zuckerberg took to the stage at Georgetown today to defend freedom of expression. Most recently, Facebook has been criticized for allowing politicians to post misleading ads. But Zuckerberg is calling for more free speech and cautioned against "potentially cracking down too much" on social networks.
Google trains its AI to accommodate speech impairments
For most users, voice assistants are helpful tools. But for the millions of people with speech impairments caused by neurological conditions, voice assistants can be yet another frustrating challenge. Google wants to change that. At its I/O developer conference today, Google revealed that it's training AI to better understand diverse speech patterns, such as impaired speech caused by brain injury or conditions like ALS.
Tim Cook calls for GDPR-style privacy laws in the US
Apple CEO and long-time data privacy advocate Tim Cook has made an impassioned speech calling for new digital privacy laws in the US. At a privacy conference in Brussels, Cook said that modern technology has resulted in a "data-industrial complex" where personal information is "weaponized against us with military efficiency," and in a way that doesn't just affect individuals but whole sections of society.
MIT's AI can tell if you're depressed from the way you talk
When it comes to identifying depression, doctors will traditionally ask patients specific questions about mood, mental illness, lifestyle and personal history, and use these answers to make a diagnosis. Now, researchers at MIT have created a model that can detect depression in people without needing responses to these specific questions, based instead on their natural conversational and writing style.
BabelOn is trying to create Photoshop for your voice
Speech synthesis -- the process of artificially creating the human voice -- isn't anything new. But a startup from San Francisco called BabelOn is working on a particularly unique offshoot of this technology. In a nutshell, BabelOn wants to make it a trivial matter to translate your own voice into another language, even if you don't speak that language yourself. The company says its combo of software and custom-built hardware can analyze what makes up your voice and then use that to recreate speech that sounds just like you, in a language of your choosing. Initially, the company wants to use its technology for things like improving dubbed films or localizing video games, but eventually it wants to be able to translate your speech in real time, say while you're on a Skype call. Microsoft has done this for a while, translating Skype voice calls on the fly, but BabelOn promises that its translations will sound like you, not an anonymous Siri- or Cortana-like digital voice.
Facebook briefly bans journalist's post slamming Trump supporters
Journalist Kevin Sessums, known for celebrity profiles and memoirs, called Trump supporters "a nasty, fascistic lot" in a Facebook post yesterday. Shortly thereafter, it was removed for violating the network's "community standards" and Sessums was blocked from posting for 24 hours. Only after being contacted by The Guardian did Facebook reinstate it and issue a mea culpa, stating the post was removed in error, but it's the latest in a year of questionable actions in which the social network temporarily locked out journalists or briefly banned content in alleged error.
Researchers say dolphins speak to each other in full sentences
It's not just the family dog that can understand complex language. Researchers at Karadag Nature Reserve in Feodosia, Ukraine have made an incredible discovery: dolphins are able to speak to one another in complete sentences containing up to five words apiece.
Dogs understand what you're saying, not just the sound of your voice
One pinnacle of human development has been the ability to comprehend language, thought to be a major achievement separating us from the animals. So it was assumed with dogs, but new research reveals that canines understand more subtleties in speech than we thought thanks to similarities in brain regions to those of humans.
Wikipedia is developing a crowdsourced speech engine
Wikipedia announced on Thursday that it is collaborating with researchers from Sweden's KTH Royal Institute of Technology to develop an open, crowdsourced speech engine that will make the online encyclopedia more accessible to people with reading or visual impairments. Wikipedia estimates that 25 percent of its user base -- approximately 125 million people monthly -- will benefit from the new service. And while the engine will be optimized for use on Wikipedia itself, any site running MediaWiki software will be able to integrate it as well.
Amazon will stream the President's State of the Union speech
The President's State of the Union address has been viewable online for years, but you've had to visit either the White House's website or YouTube. That's not very convenient in an era when streaming services reach just about every device you know. And the government knows it, apparently: the Office of Digital Strategy has announced that President Obama's final State of the Union speech will stream on Amazon Instant Video (both live and on-demand) in addition to the usual sources. The goal is to bring video to where people watch, rather than make you hunt it down. If you want to tune on your Fire TV, you can.
Windows 10's browser reportedly lets you search with your voice (update: pics)
Windows 10's oft-reported Spartan web browser may not just be a leaner, fresher substitute for Internet Explorer -- it could have a few tricks up its sleeve, too. Sources for The Verge claim that Microsoft's voice-guided Cortana assistant will be present both in the OS and in Spartan -- much like Chrome's "OK Google" feature, you can reportedly open a new browser tab and ask Cortana to look something up, whether it's a website or your flight itinerary. There's also talk of pen-based annotations for websites that you can share with others through the cloud.
Facebook just bought a speech recognition company
Facebook is clearly eager to check out new interface concepts these days. Just months after its acquisition of Oculus VR wrapped up, it's buying a speech recognition company, Wit.ai. The social network isn't saying just what it plans to do with its new purchase, but Wit.ai's focus has been on a platform for voice-guided natural language interfaces. It's not a stretch to see Facebook giving you ways to dictate your status updates or chats. Also, voice recognition is particularly important for virtual reality, where you can't always reach for a keyboard -- this may play an important role in Oculus' immersive experiences going forward.
Scientists can make your inner monologue audible
When you hear someone else speak, specific neurons in your brain fire. Brian Pasley and a bunch of his colleagues discovered this at the University of California, Berkeley. And not only that, but those neurons all appeared to be tuned to specific sound frequencies. So, Pasley had a thought: "If you're reading text in a newspaper or a book, you hear a voice in your own head," so why can't we decode that internal voice simply by monitoring brain activity. It's similar to the idea that led to the creation of BrainPort, which lets you "see" with your tongue. Your eyes, ears or vocal chords don't really do the heavy lifting, it's your brain. And if you can give the brain another source of input or output you might be able to train it to approximate a lost ability like speech.
