61 replies to this topic

[heffneil]

Several years ago I think there was a link from this website to someone's demonstration of his house where he is showing how he can talk to his computer and adjust light levels and thermostat controls. It looked pretty awesome and I don't know how real or hoax it was. Anyway he talked to the computer almost like Star Trek TNG where he would say "computer?" and the computer would answer. Anyway I am curious how he did this and a link to the video. It was pretty remarkable. If someone can send me some info on this I would greatly appreciate it!

Thanks,

Neil

[Work2Play]

It was probably HAL - you can search for that term. HAL allows for voice activation and I've read of some pretty elaborate installs.

On the newer side, someone has an example of an iPhone Siri hack to control parts of his home as well.

Hard to give better links on my phone but I'll try to dig some up later.

[Frunple]

I remember exactly the video you are referring to. The guy used Microsoft's speech recognition software and had microphones in every room. He used a button in every room to activate the "listening". It wasn't a "packaged" system, he created the whole setup. I can't find the link to it but it is from this site.
Anyway, if you need an example Homeseer has a VR system that actually works very well.

[NickTheGreat]

I've seen video's like that from Homeseer

Here's a setup video

Edited by NickTheGreat, 07 August 2012 - 08:09 AM.

[etc6849]

Premise can do this with the free speechcommander module discussed here:
http://cocoontech.co...-use-some-help/

I've never used it though. If you do research it and use it, please post your results in the Premise forum.

[edge]

I recall an older thread with Tony Stewart doing that set-up for Steve Smith who was diagnosed with Lou Gehrigs disease. Maybe that's what you're thinking of?

http://www.setnetpro...eve_smith-1.asp

[Basildane]

Eh? I did all of this myself, it is written in C, and does in fact use sound clips from Star Trek in the responses.
Years ago I purchased sound effects CD's from the original Star Trek studios and used that in the code.

I can say "computer" and it will respond with the chirps.
I can control all functions, lights, thermostats, security...
All the vocabulary are derived from the HAI Omni configuration.

It REALLY blows people away. I've demoed it for some people and they thought it was a trick, a script, and the computers weren't actually responding to me. LOL.
I took that as a compliment.

It also displays a live floor-plan of the house showing live status of all devices. Click on the garage door and it will open or close, click on a thermostat to control it, click on a room and control it's lighting... etc.

Edited by Basildane, 08 August 2012 - 09:23 AM.

[Basildane]

Oh, you may find this useful if you are planning to roll this out.

I never use the voice control part for normal daily use, only to impress friends.
It is actually more of an inconvenience than anything else. The voice response - great, status displays - great.
Voice recognition is good, but not 100% reliable yet.

I've spent a month developing the linguistic rules to try and get close to perfect, but there can still be glitches.

[joe morris]

...
Voice recognition is good, but not 100% reliable yet.

I've spent a month developing the linguistic rules to try and get close to perfect, but there can still be glitches.

I have been thinking about this for a while. I don't know if you use Chrome or not but if you do, then maybe you have tried the voice recognition offered for Google searches? It is actually very accurate. There might be a Chrome API that would allow you to use the Google voice translation engine. Just a thought.

Joe

[Basildane]

I use dotNET to handle the voice recognition, I use it at work, so I like to leverage that technology.
The Google idea is interesting. I wonder if the internet latency would be a problem.

You see, it's not just recognizing words that is the problem. When you deploy this in the real world, you will find all kinds of background noise, kids, dog barking, the blender going in the kitchen, people watching YouTube videos in the background, music playing....
These distractions cause havoc in the voice recognition system. THAT'S the big problem.

I have built some linguistic rules so that my program only responds to syntax that it is expecting.
Such as "set master bedroom lights to 75 percent".
And it doesn't listen to anything you say unless you say "computer" first, that opens up the phrases.... Just like on the Enterprise.
After you have not spoken to the computer for more than 30 seconds, you have to use "computer" again to get its attention.

I'm still nowhere close to satisfied. To really make this work will require AI that will probably be available in 10 - 15 years. You would need AI for the computer to be able to discern the speech intended for the computer vs background conversation.

[joe morris]

I just tried a small test. I was playing a movie on my laptop while speaking to Google in Chrome on the same laptop. At 1/2 volume it discerned my voice correctly. At full volume it stated it could not discerned my voice.

I would think that even at 1/2 volume, being that the speakers are within 10 inches of the microphone that this would be louder than typical background noise.

I also tried talking to the computer form 10 feet away. In front, sideways, and behind the microphone. All 3 tries worked perfectly. But, I had no background noise.

I have not programmed in over 10 years and would still be in the Basic6 days. I also don't think the lag would be an issue. The response that I am getting is less than 1 second.

Here is a quick article that I found. It is just a http post and a JSON return using a FLAC audio file. Should be pretty simple for you to test and play with. If you do, I would be interested in hearing what you find out.

http://mikepultz.com...-api-chrome-11/

Joe

[Basildane]

You know what would be REALLY freeking cool? Using Google Translate in this.
I have 3 languages spoken in my house. It would be so cool to for my app to respond to Russian commands. I cannot do that with dotNET. There is no speech library from MS to support it.
Hummm.

[joe morris]

You know what would be REALLY freeking cool? Using Google Translate in this.
I have 3 languages spoken in my house. It would be so cool to for my app to respond to Russian commands. I cannot do that with dotNET. There is no speech library from MS to support it.
Hummm.

I don't see a readily available solution to a vocal translation using Google.

Joe

[Basildane]

It's there, but it is paid only. No free access. So, forget that.

[freedom]

Hi to all, what about hardware solution? Is there any low cost?
I'm experimenting easyvr but it's limited.
Thanks