NEW YORK, 7:28 PM, SAT JUL 26 | 20 POSTS IN THE LAST 24 HOURS | tips@gizmodo.com | RSS
UK | FR | NL | IT | DE | ES | JP | AU
Posts Tagged “

Speech

speech recognition

AT&T Bringing (a Tiny, Frustrating Bit Of) Speech Recognition to the iPhone, Others

AT&T's Speech Mashups is a web-based service that will bring voice-activated search to the iPhone, as well as other Edge and 3G handsets. Instead of managing speech recognition on the actual handset, Speech Mashups sends the audio sample to the server, processes it and sends back a text transcription or command to your phone. Unfortunately for iPhone owners, this does not mean voice dialing or speech-to-text app support. Not at all. More »

voice control

Direct Voxx Muso is Natural-Speech Voice Recognition Dongle for iPod nano

There are plenty of iPod cradles that let you remote control the device, some built-in to cars, but Direct Voxx has come up with the Muso that lets you do it by voice. It's an interesting bit of kit that doesn't require training to understand you, and lets you demand particular tracks, scan through playlists, pause and resume playing music just by speaking in natural language like "play California Dreaming by the Mamas and the Papas." Check out the video to see it in action. More »

cyborg toys

Talking Twittering Teddy Bear is Nearly A Freaky Mod Too Far

Having a cyborg teddy reading out your Twitter alerts... *shiver* the idea gives me the creeps a little. But not the guys who came up with the idea over at HyHome2.0. They've even got an instructional video so you can build your own artificial-voice bear, which uses Bluetooth to get data from your PC so you can plop the toy anywhere and still keep in touch. I'm not going to be building one: I've got a vision of teddy getting bored with endless inane Twitter updates, slapping in some steel fangs in its cyber-jaw and chewing its way out of the house. [Project page via Hackaday]

darpa

50 Years of DARPA: 5 Good Inventions, 5 Lousy Ones

To commemorate the golden jubilee of America's Defense Advance Research Projects Agency—formed these 50 years ago in response to a little traveler called Sputnik—New Scientist has come up with a short list of 10 DARPA inventions: five that changed the world, and five that fell flat: More »

voicemail

PhoneTag "Hacked" For GrandCentral, Converts Voicemail to E-Mail

PhoneTag, the gang that turns voicemail into text, has "hacked" Google's GrandCentral. The service that consolidates all of your phone numbers into one is useful for people on the go, but its voicemail system is clunky, especially for those needing fast access to messages. That's what makes this hack so clever. More »

cars

Scientist Studies Cars That Speak and Listen Like KITT

KITTdash1.jpgI used to want a car just like KITT when I was a kid, though perhaps without that slightly annoying nasal voice. And now a Stanford scientist is looking at whether nattering to your vehicle may be good for you, the car, other road users and even (darnit) your insurers or advertising execs.
More »

voiceless speaking

Audeo Neckband Translates Your Thoughts into Speech, Voicelessly

The science-fiction sounding Audeo can apparently detect nerve impulses your brain sends to your vocal cords, and then translate them into meaningful electronic speech without a sound escaping from your mouth. While secret agents everywhere are presumably rejoicing at the idea, it's actually intended to help people whose disabilities mean they can't speak. Don't believe it? The designers recently demonstrated it by making a cellphone call at a Texas Instruments conference, and recorded it on video.

More »

text to speech

Aigopen Reads Books So You Don't Have to

The Brief: Hold the Aigopen up to any content in a book (text or images), and it will read out exactly what is going on, in either Chinese or English. We were amazed with the concept, clarity of audio and miniature size. As ever, there was a catch. More »

speech rec follies

Voice Activation Lameness is All Too Real


We don't care about the Ford Sync car this commercial's trying to sell you, but we find its depiction of balky voice-activated devices all too realistic. Knowingly chuckle along with us as these tools/fools who think they've mastered voice commands get their comeuppance. If this Ford Sync is like our Honda Civic Hybrid, its voice activation might still suck. It's just too noisy in a car environment to work well. Anyway, the idea of telling machines what to do and actually having them carry out your orders is compelling, but we've noticed far more situations where those systems don't want to do what you say. Maybe someday. Open the pod bay doors, Hal. [bb Gadgets]

hands free sms

In Car, Hands-Free Text Messaging Using Microsoft Sync

Microsoft has teamed up with voice recognition software firm Nuance to develop its own hands-free text software for use with Microsoft Sync. The software would allow users to dictate SMS messages and possibly emails later on down the line. This would undoubtedly result in far less cellphone-related accidents, not to mention instances where I flip off dudes in Fords. Expect the software to be available in Ford models already equipped with Sync starting in early 2008. [Tech.co.uk]

voip

ThePudding.com Phone Service Listens to Your Calls, Makes You Watch Ads

It sounds like a double-whammy of a bad idea: a free phone service that determines which ads to target to you by applying speech-recognition to all your conversations. To make things worse, the home page of ThePudding.com insults potential customers by saying it's "a breakthrough technology that makes your phone calls interesting." Hey, my phone calls are a thrill a minute. More »

talk to me

UPDATED: iPhone Speech Recognition Demo from VoiceSignal


We've heard of VoiceSignal speech recognition for lots of other phones, but now VoiceSignal sent us a video that allegedly shows it working for the first time on the iPhone. According to the guy in the clip, a couple of VoiceSignal engineers designed this app, but all we see it doing so far is controlling music on the iPhone. More »

text-to-speech

Garmin Nuvi 260 Brings Crazy Talkin' to 'Budget' GPS

Garmin rolled out its Nuvi 260 today, the company's first GPS device in its palm-sized Nuvi 200 series to have Garmin's sometimes-comical text-to-speech capabilities. We like to make fun of our Garmin text-to-speech talker; for example, when it comes up to Forest Drive, it calls it Fo. Rest. Drive. Hahaha. Anyway, text-to-speech is how Garmin differentiates its "premium" GPS systems with all those other "budget" units it sells in the Nuvi 200 series. More »

multitasking robots

Multiple Speech Recognition To Make Sex Dolls More Fun

Kyoto University and the Honda Research Institute Japan have collaborated to form the Robot Audition Project. They are at present making some very impressive ground; they have developed an advanced speech recognition technology that has the capability to understand multiple, unique speech inputs, whilst computing their meaning independently. More »

review

Vonage Visual Voice Mail Hands On (Verdict: Mixed Success)

VoIP telephone service provider Vonage just began offering Visual Voice Mail, a text transcription service that turns all of your voicemail messages into text that's immediately emailed to you. Using a combination of speech-to-text software and human transcribers, Vonage is charging 25 cents per transcription, which could end up getting expensive if you have a lot of voicemails. We gave the service a try, with inconsistent results. More »

mommy

Computer Learns Baby Talk, Pacifies Programmers

Stanford researchers have developed a program that can, at a most basic level, learn language. In studying whether the brain is hard-wired with preset sounds or if it acquires the basics of speech dynamically, a computer program was designed to learn speech. After listening to hours of Japanese and English mother-to-baby talk, the computer was able to learn the basic vowel sounds just as a baby. The computer performed so well that its accuracy was measured between 80-90%, depending on the software architecture (and whether it was "nappie time"). More »

porto-tower of babel

Star Trek Tech Arrives in VoxTec Phraselator P2

We mentioned a Star Trek-like Phraselater translation device a few years ago here on the Giz, but now it's new and improved. The Phraselator P2 from VoxTec is a bulky-looking hand-held contraption that functions like the Tower of Babel in your hands. More »

control freak

Logitech's Harmony Remotes to get Speech Recognition, Biometrics and Search Function

Just when you thought they couldn't cram any more features into their Harmony 1000, the folks at Logitech are planning to give their uber remote a trio of features that'll include speech recognition, biometric security and a built-in search function. More »