• We’re currently investigating an issue related to the forum theme and styling that is impacting page layout and visual formatting. The problem has been identified, and we are actively working on a resolution. There is no impact to user data or functionality, this is strictly a front-end display issue. We’ll post an update once the fix has been deployed. Thanks for your patience while we get this sorted.

Use Win8 Bing Voice Search feature - MS stores your voiceprint!

Of course Bing stores your voice prints. How else are they supposed to analyze them and improve their voice recognition?
 
Essentially all voice-recognition software stores this type of information to improve performance. Google even feeds similar information between services.
 
People are way to worried about this stuff... Big brother is not watching you. 🙂

I'd be more worried about little brother. The guy who takes all these voice prints and sells them. If you use voice search enough, they will have enough clips of your voice to assemble a complete voice impersonation. Imagine what you could do with the voice of someone like an Obama or a Blankfein. But really its just some random schmucks voice that could be most valuable to a criminal.

Since this is Bing, they already have all your text search records matched to an IP address. So the ability to tie that in with a voice is an added bonus. Add to that a few caches of your facebook pages and they're all set. Sociopaths are quitting their government jobs in droves and going to work in the bowels of Facebook, Google, and Bing...

Pity the fool who thinks that any of this information should ever be allowed to be collated. You just cant fix stupid I guess.
 
Looks like a ready made app to be used by Abby to solve a problem in NCIS. 🙂
 
Of course Bing stores your voice prints. How else are they supposed to analyze them and improve their voice recognition?

Why do they need to analyze anything to improve it? I paid $25 for an app called Vox Commando, it took about 2 minutes for it to analyze my XBMC music/movie libraries and it's 99% accurate with zero training for everything I throw at it. Now I gotta assume a program written by 1 dude in his basement can't possibly be anywhere close to as advanced as what MS can come up with right?

I'm not saying it's a big deal that MS does this, but is it needed? no...
 
Why do they need to analyze anything to improve it? I paid $25 for an app called Vox Commando, it took about 2 minutes for it to analyze my XBMC music/movie libraries and it's 99% accurate with zero training for everything I throw at it. Now I gotta assume a program written by 1 dude in his basement can't possibly be anywhere close to as advanced as what MS can come up with right?

I'm not saying it's a big deal that MS does this, but is it needed? no...
If I recall correctly, you're from where, California? If so, then you probably don't have an accent since for whatever reason the speech patterns of the west coast (particularly the Pac NW) are neutral by English standards. Whereas if you had an accent, then the software would be far less accurate if it couldn't learn from its mistakes. There are a lot of people with a lot of accents out there, and the only way to improve is through analysis and machine learning - it's not something you can program for.
 
If I recall correctly, you're from where, California? If so, then you probably don't have an accent since for whatever reason the speech patterns of the west coast (particularly the Pac NW) are neutral by English standards. Whereas if you had an accent, then the software would be far less accurate if it couldn't learn from its mistakes. There are a lot of people with a lot of accents out there, and the only way to improve is through analysis and machine learning - it's not something you can program for.
Lol, you all talk funny because TV people talk like us in Los Angeles.

Until i was about 10 years old, I didn't understand that all police cars weren't black and white. They were black and white in my town, and on TV, so they must be all black and white, right? 😉
 
Idiomatic expressions also are telling. In the NYC area, people stand on line. In the west, they stand in line - accent or no accent. 🙂
 
If I recall correctly, you're from where, California? If so, then you probably don't have an accent since for whatever reason the speech patterns of the west coast (particularly the Pac NW) are neutral by English standards. Whereas if you had an accent, then the software would be far less accurate if it couldn't learn from its mistakes. There are a lot of people with a lot of accents out there, and the only way to improve is through analysis and machine learning - it's not something you can program for.

Great point, one I hadn't though of. Bing won't be able to understand a Bostonian who says "find closest Pizer joint" I guess voice recognition needs a lot of help to understand those weird back east accents.
 
I'd be more worried about little brother. The guy who takes all these voice prints and sells them. If you use voice search enough, they will have enough clips of your voice to assemble a complete voice impersonation. Imagine what you could do with the voice of someone like an Obama or a Blankfein. But really its just some random schmucks voice that could be most valuable to a criminal.

Since this is Bing, they already have all your text search records matched to an IP address. So the ability to tie that in with a voice is an added bonus. Add to that a few caches of your facebook pages and they're all set. Sociopaths are quitting their government jobs in droves and going to work in the bowels of Facebook, Google, and Bing...

Pity the fool who thinks that any of this information should ever be allowed to be collated. You just cant fix stupid I guess.

This entire post cracked me up. Bravo sir. If somebody wanted Obama's voice, they could....oh, I don't know.....maybe go through the hundred's of hours of him speaking on tape? The fact that you even think somebody has the technology to take all these samples and put them together to make an even somewhat believable voice to interact with another human is laughable.
 
This entire post cracked me up. Bravo sir. If somebody wanted Obama's voice, they could....oh, I don't know.....maybe go through the hundred's of hours of him speaking on tape? The fact that you even think somebody has the technology to take all these samples and put them together to make an even somewhat believable voice to interact with another human is laughable.

There actually was an app for the iphone, where you could type in phrases, and it would speak them with the voice of a famous person or president. They did go through many recorded copies of their voice, to extract phonems, and splice them together.
 
So you believe somehow, someone will hatch some evil plan using your voice prints that will hurt you in some crazy way? I'd much rather be ignorant than live in a world of a delusional paranoia.

Did I say that? You did. I believe there's a copious amount of surveillance in this day and age, and a loss of privacy. That's not delusional paranoia, it's called opening your eyes.

I'm not a conspiracy theorist but I'm also not a coincidence theorist.

I don't want to derail the thread.
 
Last edited:
don't the voice recognitions in Apple and ANdroid also do that? It's supposed to learn how you speak

I speak with a foreign accent. And at first my Android phone could not understand me no matter what I did. Every time I tried voice it came out botched and totally unrelated to what I was saying. I thought it was silly considering they had Maria Sharapova talking to the thing in Russian!

Now after using it probably once every two weeks I've noticed it's more accurate and gets what I mean from the first or second try unlike before when it sometimes thought I was saying boobs instead of boots 🙂
 
don't the voice recognitions in Apple and ANdroid also do that? It's supposed to learn how you speak

I speak with a foreign accent. And at first my Android phone could not understand me no matter what I did. Every time I tried voice it came out botched and totally unrelated to what I was saying. I thought it was silly considering they had Maria Sharapova talking to the thing in Russian!

Now after using it probably once every two weeks I've noticed it's more accurate and gets what I mean from the first or second try unlike before when it sometimes thought I was saying boobs instead of boots 🙂

I don't know about iOS, but in Android it's optional. If you go into the Voice Search settings there's a checkbox for "Personalized Recognition". And if you change your mind, you can go into your Google Dashboard and remove the recordings.
 
Back
Top