It’s been almost a year since Google introduced a feature that allows iPhone owners to search the Web using voice commands, and now Microsoft is getting into the game, too. As announced on the Bing Search blog, Sprint Wireless’ brand new Samsung Intrepid phone now comes equipped with a fresh voice user interface from Tellme, a speech-recognition company that was acquired by MS in early 2007.
As the video below demonstrates, you can use the new interface to search the Web by speaking your search query, compose a text message or dial a contact by simply talking instead of typing. The company cites a study from Sanderson Studio that found 40% of smartphone usage occurs in multitasking scenarios where the user cannot offer their undivided attention to their phone, and claims the new voice command feature should simplify their lives.
Despite what the title suggests, I think it’s perfectly fine for Microsoft to add features that plenty of people will find useful, whether Google came out with them ages ago or not. Competing companies copy stuff from each other all the time, so it’s hardly something to get overly worked up about.
Of course, the feature is restricted to just one phone on one carrier only for the moment, so in the meantime you can check out services like Dial2Do, although that application is more suited for action commands than searching the Web.
I hope in time Microsoft comes out with apps for the most popular platforms rather than keeping this type of feature phone-specific.
This is awesome!. I had something similar when I used to have palm treos. Can’t wait for them to make it available on HTC phones. I want to use it on my HTC Touch pro 2 !
You don’t need to wait. Windows Livesearch has had speech recognition for a couple years now (before google was doing SR on web searches), they just rebranded to Bing but it’s the same service.
It doesn’t do text meassages but it does location based voice searches for most things you need from a mobile device. Cheapest nearby gas, restaraunts+services, etc.
Hey, everything is going to be voice operated soon. In the not too distant future, keyboards will be optional and only needed for detailed and work specific tasks I.E. Photoshop, gaming, etc. It wouldn’t be surprising if Google is already working on telepathic control!
Doubtful.
that wont happen.
what if I am the following things
-mute
-unable to speak english
-heavy accent
-have a low register/diction/vocabulary
-mute
But people who are incapable of using their hands will be able to use it (or with one hand, or with only fingers lost), so that’s always a trade off
-Such systems will over time become perfectly multilinqual
-I have an extreme accent and most speech to voice systems fail for me, but e.g. goog411 always understands me correctly, so it is already possible
-People with a low vocabulary will have just as much trouble using computers now and in the future.
I love reading these exchanges where neither party has a clue whatsoever. You guys, you crack me up.
n00bs
I like your optimism.
Running off to work so don’t have enough time to write a full response but:
1. Hasn’t Windows Mobile already had Voice Command and Voice Search in the Bing Mobile app? I saw the video and noticed that this is now using the TellMe service that they purchased a while ago, but otherwise they’ve had voice search and command capabilities for several years. This just looks like the voice search capabilities are much more integrated and there’s a speech-to-text capability.
2. Who really uses voice commands and search? I don’t think there’s a big market for this. Rather I think it’s more of a gimmicky feature. I’ve had a Windows Mobile phone for several years with Voice Command and barely ever use it. I’ve used the Voice Search of Bing Mobile but even then it doesn’t really provide any time saving.
If anyone really uses the voice search/command features, I’d love to know when and where it can really be useful.
I’m a fast PC keyboard typist and therefore go bonkers when trying to type using a smart phone keypad, hardware or software (Android G1 or iPhone). Google’s Voice search is accurate enough that for a (whole) lot of searches it accurately translates your voice query to text. Some queries I have done are “museum of natural history”, “Domino’s pizza”, names of local cinemas, etc. That’s a *lot* of typing eliminated.
In addition, If TTS ever gets fully utilized in more than driving apps (Android now has TTS with the release of 1.5 aka Donut ), you could have apps which you could use while driving since you can listen to the results of a voice search instead of trying to read the screen which is dangerous. Of course they’ll have to get rid of the need to touch a button to start the voice search or make the voice search trigger something you can “feel out” with your fingers without looking.
Well said Robert. Voice activation eliminates a lot of the dangers associated with looking at your phone when we should be looking at the road. That applies to all interactions with your phone, whether making a call or opening up sat nav. There is a huge market for this, the more integrated it becomes the more usable it becomes, the more usable, the more scalable. PS – Typing on a BB Curve or the (very very similar looking Samsung) is the easiest phone on the market to type into, even without looking at the phone. New smart phones mean we need new technologies like these to compensate for no raised buttons. Look forward to seeing this on BB.
I meant version 1.6 for Donut.
I’m a windows user.I like windows,but bing is ok.I love google.It is really good and simple.
Thank you for your contribution to this discussion
Mark just provided me with a profound fact
Ha ha Robin – It makes so much sense that this is the comment you would reply to rather than acknowledge your post’s poor research as pointed out by nearly everyone else. It would be one thing if we could assume you didn’t even read the comments but there you are!
Funny joke, terrible job.
I think the voice recognition software’s still need some ironing out. I have G1 and I have tried it numerous times and it still doesn’t seem to work like it should. Good step in the right direction. I think just a bit more time and research is needed.
I agree that it is short sighted to make these aps phone specific. Technology is constantly changing so they should be more universal.
1, Microsoft might bring out a lot of new seemingly scattered features for bing over a median period of time. Since Google’s development process is screwed up, so that will put a lot of pressure on Google. And it keeps them in the news.
2. Voice recognition is labeled right for once.
Recognition != Understanding
In other words for index search for some cases it can be good enough.
In general, voice as an interaction medium is excellent. Voice as a one way input system is:
Let’s just say not very smart.
I’ve been using voice search on my Windows Mobile phone for over a year, maybe two. Windows Mobile Live Search had it as did the rebranded “Bing”. It also returns web results. Perhaps they are simply advertising a capability they already had or expanding it further? Either way I’ve enjoyed using it for quitte some time and was impressed by how well it worked.
Google’s work on voice search is terrific, but if anything it was inspired by Tellme, not the other way around.
Mike Arrington covered an earlier version of the Tellme voice mobile search product in 2007 (”Tellme Launches a Killer Mobile Product”):
http://www.tech...obile-launches/
Note: I am biased
(co-founded tellme, left earlier this year)
That said, in mobile, being first and/or better doesn’t matter without distribution, and distribution decisions in big companies are sometimes complicated.
Hats off to the Tellme team for launching operator-free text message transcription. Not an easy feat!
copy copcy copy copy….just like MS – “let’s see what everyone else is doing – then we go buy a company and badly implement the same concept so that we can play too” !
god shut up. gmail, google maps, iGoogle, google docs, google checkout, whatever their social network was called …. all copied.
I think what we see in the comments is People thinking voice input equals writing.
Well, it’s all about feedback loops coupled with input.
If we are writing, we are used to getting feedback on our spelling, from the point in time we learned writing.
If we are talking we get feedback sometimes on pronunciation (younger) as we grow old we get feedback on context. So feedback on spelling while dictating even it’s the “damn” systems fault is to say at least irritating to us.
Even if the system recognizes all words correctly, hard to do since there is no understanding and human language is not exactly precise , it will feel in most cases weird. Since we are expecting feedback on context. Or try to explain something to a person behind a black screen with no feedback except pronunciation corrections …. don’t hurt him/her.
Yeah, they’ve had voice recognition in the Live Search/Bing App for WM for nearly 2 years now, I think…. This is just a new implementation for 6.5 that also lets you transcribe text messages, call people, etc.
I think it is great idea to have this feature, but there is still a lot of work needed to improve on the feature. I think it will get better in the future.
And in spite of numerous people already pointing out that they’ve had this feature available for nearly two years now TechCrunch has yet to correct their report’s title.
It’s almost as if I am getting my tech news from a news programme or something, what a joke TechCrunch you should either be more informed to start with or at least do more research.
Robin Wauters, I appreciate the google-bias, but as other readers have pointed out, Microsoft has had speech recognition long before google – live search has had it since 2007. they’ve simply rebranded.
perhaps an edit is in order?
Google, Yahoo and Microsoft all have this and it is equally useless for them all. Voice input for arbitrary strings is just not there yet. Especially when you are in anything but a silent environment.
This just does not work. But Bing has to have it, to keep up.
Like others have said, this feature has been part of the live download apps for most mobiles for a couple of years. I think the new Bing brand is new. I had this on my Samsung Win Mob nearly 3 years ago.
You should update the article.
“It’s been almost a year since Google introduced a feature that allows iPhone owners to search the Web using voice commands, and now Microsoft is getting into the game, too.”
I’ve had this on my Windows Mobile Samsung Blackjack II long before Google “got into the game”. At the time it was called Live Search but doesn’t make any difference. Get the facts straight.
Same here. My Blackjack II has had the voice recognition for the Live Search app for years, and it’s been great. I’m surprised this story hasn’t been updated with that clarification yet, given it’s been published for several hours now. The implication in the article (as well as the title) is that voice recognition new to Microsoft’s mobile app for search and is just now ‘getting into the game’. This has been thoroughly discredited, so why no update?
Tellme is a very powerfull voice application also in the automotive industry.
But this version seen on the video uses some wasted machine cycles, I mean come on do you need to watch the screen that its
( Listning , Thinking or what else)
Thats crap and needs to be cleaned up. It gives a feeling of having a toy on your phone its a smartphone not furbie .
I got something similar on my asus smartphone it does what it supposed to and not tell me it does.
Waste of battery power my other point.
It’s no wonder Google and Apple get so much love when there’s this much clueless knob-polishing going on by high profile tech news sites feeding the ignorance.
I’ve been using voice search on Live (now Bing) search for 2 years now (long before Google had it.) It’s very useful when you’re in the car, etc (although I’d argue Bing 411 and Goog 411 are more useful . . . and I’ve found the former to actually be a bit better than the latter as it has a more robust featureset.)
And the TellMe service is NOT just going to be phone-specific, but will eventually be available on most Windows phones (and possibly could roll over to other platforms as well, I know Live/Bing has been available on Blackberry for awhile now.)
How is this “inspired by google”? They invented voice control? Such a dumbass comment.
MS has versions of this on their phone for YEARS. Google has been doing it for about a year, as you say. How come when they introduced it it was not “inspired by MS”? ou idiots don’t know anything exists until Apple or Google does it – then you are blown away. You’d be fine if you didn’t have to get in an MS dig but then, what’s the posting right?
Fix the headline – your ignorance is an embarrassment to TC.
There are many flaws in the article but given the author is not in this particular industry, many are understandable.
First, Tellme is not and never was a speech recognition company. They had very little speech technology but rather licensed almost all of it from Nuance (and later IBM and then Microsoft’s after the acquisition). Rather, they were primarily an out-sourced IVR hosting company. They had their own voice portal and got into mobile-specific apps beginning in 2005. They released a downloadable (RIM platform) speech-enabled local business search application in 2006.
In 2008 (specifically the CTIA show in Vegas in April ‘08), Yahoo released the first wide-open speech-enabled web search application. This was called, cleverly enough, Yahoo OneSearch with Voice. In the small speech universe this was a breakthrough application. In less than a year Google responded with their own wide-open speech-enabled web search application. MS followed some time later.
Speaker-independent speech reco in the cloud is a very difficult task. As legislation mandates more speech and less screen/keyboard usage in cars, expect to see more resources thrown at the problem. Network bandwidth, microphone placement and pattern on the device, network coverage…all conspire to work against effective recognition. Expect to see evolutionary rather than revolutionary advances in the very short term.
*applause* someone who actually knows something. hurrah!
I’ve had this feature built into my phone since Windows Mobile 6.1.And it’s not only search. I can for example say “what’s my schedule like today?” and it reads me back what’s in my calendar.
And Live Search on Windows Mobile had it even before that. So to be correct Microsoft introduced it at least 1-2 years before Google.
There seems to be a lot of confusion as to the “feature” we’re addressing. I believe you think it’s speech recognition…in your case, embedded recognition. That’s not what the article was discussing. Embedded reco has been around for years. We’re discussing speech recognition in the cloud for unbounded web search and short messaging. This is a very different, far more complex problem. Embedded recognition, the type you’re describing, is recognizing a few keywords within a very restricted grammar. This is all done on device. Google does not, to my knowledge, do any embedded recognition. The embedded reco found on Android phones is not theirs.
this just proves that technology didn’t exist until the iPhone was released.
To state that Google somehow invented mobile speech recognition (SR) web search and that Microsoft is copying these features is flat out wrong. If anyone had it first it was Microsoft. Their Live Search app with SR web search was out a long time before Google ever even though about writing an app with that functionality. Of course since you get a big fat check from Google every month from your advertizing, I’m not really shocked to hear it spun this way.
We’ve actually been working to enable voice commands on mobile phones since our original Voice Command client for Windows Mobile debuted in 2003. One of my good friends, Robert Rebholz, led the charge on this and would wax poetic for hours about the utility of being truly hands-free to accomplish tasks while mobile. I’m also pretty sure we were the first to add voice search to our then Live Search client for Windows Mobile back in Oct 2007. But it’s less about who was first and more about how great it is to see more players innovating in this space.
I like a lot of the comments I see in the stream here and agree with a good number of them. There are a ton of challenges in making good voice search (a decent signal acquisition, compression, client-side processing power, correction, etc) and so we see voice as one of the many modes in which you will interact with a search engine. This advance in voice search and human-computer interaction really speak to the larger Bing strategy – that search needs to better adapt to user intent – both in how we let you query and in how we display the results that make the most sense for the given situation. Search is becoming more embedded in everything we do – the white box and ten links is still very useful for certain types of queries, but as people query for less definitive concepts (where should I go on vacation this year?) the engine needs to adapt its response. Whether that means a more visual experience, pre-processing of large data sets to provide insight rather than just information, or a guided model that can help you through exploratory queries – search will become the universal agent that connects your virtual to physical world.
To that end, a great voice search is yet another method by which we can make search more natural and less a ‘task you do’ and more like looking at your watch to tell the time. It’s natural, automatic, and non-intrusive. Stay tuned for more!
Actually, the cloud reco added to LiveSearch in 2007 was strictly local business search. Tellme had deep expertise and experience in Directory Assistance given their work with Verizon landline and Cingular (AWS) mobile 411. Tellme leveraged this expertise and created the first multimodal business search client. The first unbounded speech web client was done by Yahoo in conjunction with vlingo. It was unveiled in early 2008. They were followed by Google and then Microsoft/Tellme. I worked for both Tellme and Yahoo and oversaw the development of both clients. Of the three, only Microsoft had expertise in embedded. Nuance has a great deal of expertise in both embedded and cloud reco…perhaps more than anybody.
Microsoft was also inspired by Google’s and Yahoo’s funny names so instead of using a normal name like Live.com, they used Bing. Why use proper English when you can copy other companies’ fabulous ideas?