http://www.iflscience.com/technology/it-s-not-just-your-tv-listening-your-conversation
Stop talkig too loudly – Internet can hear what you say.
You might not be too paranoid if you are worried that your new TV and smartphone can listen every word you say so that they can send information on it to Internet – because they can do that in many cases.
Posted from WordPress for Android
3 Comments
Tomi Engdahl says:
The Ear of Sauron
https://www.f-secure.com/weblog/
A recent story by The Daily Beast seems to have ignited a real firestorm over Samsung’s “smart” television terms and conditions. Which is somewhat surprising to us as we read about it months ago via Mikko. But anyway, things that listen are topical.
So… do the words “always-listening voice search” sound good to you? Or do they give you the creeps?
Because that’s the potential future of Google’s Chrome browser:
The “always-listening” feature is currently available via: Google Voice Search Hotword (Beta)
That’s the thing about voice “activated” devices. They’re always listening. Always recording (to a buffer). The question is: how much gets uploaded to the voice recognition service?
Are you comfortable with a “few” seconds?
Tomi Engdahl says:
Your Samsung SmartTV Is Spying on You, Basically
http://www.thedailybeast.com/articles/2015/02/05/your-samsung-smarttv-is-spying-on-you-basically.html
You may be loving your new Internet-connected television and its convenient voice-command feature—but did you know it’s recording everything you say and sending it to a third party?
Careful what you say around your TV. It may be listening. And blabbing.
“Please be aware that if your spoken words include personal or other sensitive information, that information will be among the data captured and transmitted to a third party,” the policy reads.
So be advised: If you’re too lazy to pick up the remote, you may want to keep your conversation with the TV as direct and non-incriminating as possible. Don’t talk about tax evasion, drug use. And definitely don’t try out your Violet Crawley impression.
But, said McSherry, “If I were the customer, I might like to know who that third party was, and I’d definitely like to know whether my words were being transmitted in a secure form.” If the transmission is not encrypted, a SmartHacker could conceivably turn your TV into an eavesdropping device.
“Samsung may just be giving itself some wiggle room as the service evolves, but that language could be interpreted pretty broadly.”
Tomi Engdahl says:
Voice as an interface in the smart home: Can you hear me now?
http://www.edn.com/design/consumer/4438627/Voice-as-an-interface-in-the-smart-home–Can-you-hear-me-now-?_mc=NL_EDN_EDT_EDN_consumerelectronics_20150211&cid=NL_EDN_EDT_EDN_consumerelectronics_20150211&elq=6e68789c488d47e682933b3a60326942&elqCampaignId=21576
It is not difficult to imagine a future where humans speak to their smart home and have as meaningful an interaction as they do with each other. Such ideas have been portrayed in television shows and movies even as early as the 1960s.
However, there is an economic challenge to building a practical product with good performance. As most end consumers don’t have the resources of Tony Stark, the product should be able to provide excellent speech recognition performance within a limited cost budget. In particular, traditional beamforming voice processing technologies require many microphones for typical smart home applications
There are many technical challenges in developing such a system
This discussion will focus on an often overlooked technical challenge in far-field speech interface systems, yet one that is just as critical: pre-processing of the speech before it reaches the speech recognition engine.
Even the most modern speech recognition engines have one basic requirement to work well – the input to the engine should be speech. Although this may seem obvious, for far-field speech interface systems, it is one of the most significant challenges. In this case, “far-field” refers to a system where the user’s voice is more than half a meter away from the product’s microphones.
Implementing a system that enables robust speech pickup in these conditions is a challenging task.
This article will show why conventional approaches fall short of delivering acceptable performance under these far-field conditions, then propose a solution that provides good far-field performance in a cost effective manner.