Now that Google has released Voice Search for it’s Google Mobile App for iPhone, reports of problems with the speech recognition are being reported.

Evidently, Voice Search works really well in North America, but is posing some challenges for users Britain.

For example, one user spoke a search for “fish,” only to retrieve the search results for “sex.”

Of course, we at Speech Tech Blog can relate.  Every time I do any type of web search, I only get search results about sex.  But maybe that’s just me.

Below are a number of stories about similar problems with Voice Search:

Link 1

Link 2

Link 3

Link 4

As promised, the Speech Technology News Feature about Google’s new voice search for iPhone.

A few days ago, we brought you this post about Google’s new voice search for the iPhone.

And as expected, Google is being typically reticent about this development.

But fear not, gentle reader: We at Speech Tech Blog are not done yet.

In fact, as soon as I am done with this blog post, I have a strongly worded email to send to a certain massively-powerful-and-ridiculously-secretive-company.  A company that has access to all my personal information; a company that has a record of every Internet search I have ever made; a company that will, for obvious reasons. remain nameless.

But in the meantime, check out the following links for more information on Google’s voice search for iPhone.

Official Google Blog

Official Google Mobile Blog

Google Press Center

According to a story in today’s New York Times, Google researchers have added voice recognition to their search software for the iPhone.

According to The Times, the new application will be free and available via the iTunes store.  Basically, users will be able to simply speak searches into their iPhone.  The results will then be displayed on their phone and include local information when applicable via iPhone features that detrmine user location.

We expect an announcement about this from Google, so keep your eyes peeled for a Speech Technology News Feature in the near future.

In the meantime, John Markoff’s New York Times story can be found here.

So, I came across this strange and oddly worded press release today, which announces the creation of a toolbar for Internet Explorer that enables speech recognition for Google’s search engine.

According to the release, when the program–created by developers from the Ukraine–is installed, users may enter a search by voice and then either click the search button or say “search!” This redirects users to www.voicesearchbar.com where they will see the results of their search.

From the release: “Authors of the program recommend launching the training mode to enhance the quality of speech recognition. After 3-4 launches of a training mode program recognizes speech pretty good.”

Hmmm.

Apparently, the VoiceSearchBar toolbar is available for download at www.voicesearchbar.com.

I visited the site–it features an attractive blond woman (who evidently likes speech enabled internet searches) and comes with a special dedication: “Dedicated to Bill Gates: ‘In five years, Microsoft expects more Internet searches to be done through speech than through typing on a keyboard.’”

Hmmm.

The site also features an About page that allow users to make donations.

Hmmm.

I really don’t know what to make of this. Part of me feels like if I were to download VoiceSearchBar, my bank account would be instantaneously drained, a vast array of pornographic images would be sent to everyone in my email address book and my computer would explode.

In search of a second opinion and in an effort to not malign what could be a perfectly legitimate product, I contacted a friend in the Speech Tech World.  He seemed to think that my instincts were correct.  He also thought it odd that VoiceSearchBar was being released for Internet Explorer, which is not an open source browser.

I don’t know: Maybe VoiceSearchBar is totally legitimate.  Maybe not.  Does anyone out there have any information about this?

Last week, Google implemented optional audio previews for YouTube comments that allow users to listen to comments before posting them.

What made this move somewhat buzzworthy was that it came shortly after Randall Munroe’s popular webcomic XKCD suggested that Google add audio preview for all comments–so that people might realize how inane the majority of their comments really are.

And really, you can’t argue with that.  A random visit to YouTube yields the following:

The YouTube “Spotlight” was “Project’s Slidy Interlude.”  I clicked on it and watched “Interlude: Slidy“–which was actually pretty cool: a beatboxing flutist with string instrumentation.  But let’s look at the comments.  Here are the first three:

• sweetcandy719: “anyway, Lo and behold, I found my man cheating on me with some whorish babe named Julia…”
• mystikcateyez: “Pretty cool music. Stop by my channel and check it out. Thanks alot for making youtube worth watching.”
• godbil4: “awsome”

Truth be told, it could be a lot worse.  I have to say that sweetcandy719 probably could have benefited from hearing her comment read back to her.  But, that is neither here nor there, I suppose.

My next move was to contact Google for some comment on Randall Monroe and this new speech feature.  My first email was met with stony silence.  The following email–sent a day later–yielded a response: I was told that my message would be passed along to someone at YouTube who would get back to me.

Someone at YouTube never got back to me.  Luckily a lifetime of rejection has given me thick skin and I only wept for a few hours.

But back to the audio preview.  I decided to test it out with a random comment for “Interlude: Slidy.”

When asked to speak my comment–”And they have no disregard for human life.” ~ George W. Bush, on the brutality of Afghan fighters, Washington, D.C., July 15, 2008–the speech technology worked pretty well.  It didn’t read the quotation marks and had a little bit of trouble with the president’s middle initial, saying “west” instead of “W.”  But, all in all, not too shabby.

So, whom do we have to thank for this new feature?  Google?  YouTube?  Randall Monroe?

The world may never know …

As Speech Technology reported here a few weeks ago, Google launched a new audio search indexing experiment that allows users to find spoken words inside videos.

Google Audio Indexing (GAUDI) was developed by Google Labs and lets users search for spoken words inside video clips and jump to portions of a video where the searched words are spoken. For now, the GAUDI tool is available only for election videos, but Google plans to expand its use to other videos.

More importantly, GAUDI works well and is pretty fun to use. So, if you want to locate particularly rousing sections of John McCain’s convention speech based on key words or relive Sarah Palin’s sage remarks about lipstick and pit bulls, GAUDI may be the tool for you!

This is a transcript of the morning keynote panel between

Bill Meisel, President, TMA Associates

Mike Cohen, Manager, Speech Technology Group, Google

Victor Melfi, Chief Strategy Officer & Senior Vice President, Marketing, VoiceBox Technologies

Neal Bernstein, Senior Director, Local & Mobile Search, Microsoft

Michael Wehrs, Vice President, Evangelism & Industry Affairs [Ed: Yes, that is his real title], Nuance Communications

John Tadlock, Lead Technical Architect, Consumer Application Architecture, AT&T

I typed the discussion out as I listened. So if it’s a little sloppy in parts…tough tamales, readers.

I did this for two reasons. First, it was a really good debate and the panelists gave some great information that I was unable to include in the daily Speech Tech news story. I just didn’t have enough space.

The second reason is because one panelist calls another an “ignorant slut” [Ed: According to staffers here, this phrase originates from SNL's Weekend Update editions with Jane Curtin...and is hilarious in the context of voice search.]. I wanted to share that.

And yes. Yes. You’re welcome.

Meisel: Voice search is vague and we’ve kept it vague for this conference. Voice search is a way of implying, like web search, that you can get things quickly and easily. How do you see this paradigm? What can we do with it?

Answers after the jump!

(more…)

I’m not sure about the rest of our readership, but it’s a news-rich day here in the Speech Technology offices — early versions of the Android phone were unveiled today! Ryan wrote a brief news article about the open-source phones back in November 2007, and both he and I have been eagerly awaiting demos of the Android since.

And today, we finally got tangible evidence that the Android is progressing nicely. Just 45 minutes ago, the BBC News reported that several companies showcased prototypes at the Mobile World Congress in Barcelona. The best part? The first Android-enabled phone is, according to the article, expected to be released later this year.

Approximately one dozen companies took part in the demonstration, all of which are directly involved in the Android project. Companies mentioned included: Texas Instruments, ARM, Qualcomm, Marvell, NEC, and SET Microelectronics. Ian Drew, from the company ARM, told a BBC reporter,

What we are demonstrating on the Android platform is maps, browser, camera applications, multimedia, e-mail, and calendar - basically everything you’d expect on a mobile phone.

Of course, the partner we’re most interested in is Nuance Communications - which is, as Ryan reported, part of the project contribution team. How will speech play a role in the Android? We’re hoping it’s something similar to the mobile search/SMS dictation currently provided on the Palm Centro (read more about that here). We’ll keep you updated about Android as we learn more in coming days, weeks, or months.

Related Stories:

- Google Android phones make debut [BBC News]

- Google Android coverage [BBC News]

(Image courtesy of BBC News)