Advantages of Built in Voice Recognition vs Alexa or google home

[jjohnston]

I wanted to see what everyone's thoughts are regarding the voice recognition built into homeseer vs using an alexa or google home linked. Are there things you can do with the built in that you can't with Alexa, or vice versa? I found I can get an array mic etc, and do my own for about $20 more than the alexa (I could get a lower end array mic with only 2 microphones for less than alexa but need to test it all). Can't decide which way to go.

Any thoughts as to advantages and disadvantages to both are greatly appreciated!

[rmasonjr]

I tried the HST VR for a number of years. Lots of training and you had to manually invoke. Alexa and GH changed the landscape for me with VR. Not only with lighting control, but with thermostat control - being able to say "Alexa, what is the temperature in the Living Room" and then countering with "Alexa, set the Living Room Temperature to 68 degrees"
You would find that with the HST VR, this would be next to impossible.

[jjohnston]

Thanks! Keep the thoughts coming!

My thoughts are this.

Alexa/google home have web searches built in. So I can ask "who won the football game yesterday" or "how tall is the tallest building" and it tells me.

VR it all stays on network, so if the internet goes down, it still works
I can choose my own name so instead of Alexa or google I can say whatever name I give it...

are there things that Alexa/gh can't integrate with in homeseer? like other things vr can do that gh/alexa can't?

[Pete]

Here still using Microsoft SAPI VR / TTS. Mostly due to a collection of voice fonts over the years.

That said just reinstalled Windows 10 Pro on a Lenova Laptop using Cortana for installation and I was pleased with the VR as it worked great before connecting to the internet.

Currently the only integration to Cortana / Kinect is via a 3rd party Homeseer 3 application. It works fine.

Note here also utilize Alexa in an experimental mode.

[jjohnston]

Thanks!

I currently have an alexa at home, I have her turn on lights, etc. seems to work. I had to do a workaround to get her to turn on the TV (I made the TV a light inside homeseer).

But would the built in VR allow me to not have to put workaround type things in like this?

[Timon]

By far the biggest advantage of having built-in voice vrs. Alexa or Google is that it works when the internet is down.

There is a board available, Matrix Voice @ $55, that can do good local voice recognition, and TTS that could be programmed to work with HS if we had a way to interface it. The best way would be if we could use the Speaker Protocol but that's not open to any developers.

It could also be an Alexa as well, and at the same time, but not with all of the same features as Amazon hasn't released that ability to developers yet.

If HS would open this up I'm sure there are some here that would love to develop for the Matrix Voice board.

[jjohnston]

Thanks, I'm looking at the Respeaker. Looks like it'll link to a computer via usb and be almost like another sound card. My hope is, I change the Bin file of the Respeaker, give it a name like LR for living room, then homeseer sees them all as separate sound cards, and I can have homeseer take and send the audio based on this location. I figured with separate names, I may not need to worry about the computer not linking them up in the same order each time...

[Fellhahn]

Is there a way to use Alexa as a speak device? Or have her announce things on event triggers?

[Rupp]

Is there a way to use Alexa as a speak device? Or have her announce things on event triggers?

Not really. I think someone hacked a way but there is no native way. Hopefully in the future.

[Pete]

I am thinking in the near future we will see a "Cortana" internet voice which will be integrated to Microsoft SAPI which today is already integrated with Homeseer 3 for Windows.

Windows 10 SAPI VR/TTS is way better and faster these days versus the legacy generations of Microsoft SAPI.

Last install of Windows 10 Redstone edition I used Cortana (without any training) and it worked very nicely such that I installed Windows 10 on 3 of my laptops and two new desktops here. (well doing a dual boot configuration with Ubuntu 18.04).

Way back saw the seeds of this and relating to Microsoft automation when testing Spud's Kinect / speech integration stuff.

[AshaiRey]

All good and well but the wife here doesn't like to speak in English to get the lights on. So until Cortana can do that in Dutch it's a loss battle overhere.

[Pete]

Googling some found this...maybe it will work for you.

Cortana is available in these regions for these languages:

• Australia: English
  
• Brazil: Portuguese
  
• Canada: English/French
  
• China: Chinese (Simplified)
  
• France: French
  
• Germany: German
  
• India: English
  
• Italy: Italian
  
• Japan: Japanese
  
• Mexico: Spanish
  
• Spain: Spanish
  
• United Kingdom: English
  
• United States: English

Cortana's now on Windows 10 but only in a select number of countries. Why? Likely cause Cortana's text-to-speech (TTS) is region specific and maps to the speech patterns, idioms, and expressive style of each country. As a result, each of these voice fonts are specific to each region.

1 - Microsoft Speech Platform SDK Language support.

2 - Unlocking Cortana's Voice (Eva) for use in a UWP app

3 - Change Eva to Dutch - Netherlands nl-NL 0x0813

[jjohnston]

That brings up a question. Does the VR take place on the speaker client PC or the server? So if my server that Homeseer runs on is a win 7 computer, but the computer I have a speaker client running on with microphone etc is windows 10, does it use the windows 10 computer to determin the speech or would it use the win7 server?

[sirmeili]

I wanted to see what everyone's thoughts are regarding the voice recognition built into homeseer vs using an alexa or google home linked. Are there things you can do with the built in that you can't with Alexa, or vice versa? I found I can get an array mic etc, and do my own for about $20 more than the alexa (I could get a lower end array mic with only 2 microphones for less than alexa but need to test it all). Can't decide which way to go.

Any thoughts as to advantages and disadvantages to both are greatly appreciated!

The biggest difference is that when I tried to get local voice control working and I got an error, I was told by HST support to just use google or alexa and they didn't help me fix it.


So, my take away was that if it works, you have the benefits mentioned here, but if you have any issues setting it up, you won't get much help getting it fixed.

I'm not necessarily faulting HST for this approach, but it's a feature of their software and I was a bit upset that they didn't seem to care to help me get it working.

[Pete]

Does the VR take place on the speaker client PC or the server?

TTS / VR using the speaker client is using Microsoft SAPI on the client PC. The speaker client PC can run on any windows computer and doesn't have to run on the Window Homeseer box. On the Linux HS3 box HS3 speech talks to Linux TTS and or Windows TTS. When it is running and looking at the speaker clients it shows only the Windows speaker clients (on the LAN and on the Virtual box).

Microsoft SAPI was kludgy and slow training on XP, Vista, W7 (started to get better), Windows 8, Windows 8.1 and now really good on Windows 10 way long time ago before smart phones...and a GPRS or 2G or 3G mobile connection using speech sucked until 4G (LTE) came along. Here locally I complained to T-Mobile about LTE piss poor quality...opening a ticket with them...response and closing ticket was that the towers were saturated (?)...testing last week and today now see 35 up and down where as just a few weeks ago it was 15 down and maybe 6 up.

A while back tested running HSTouch, HS Kinect, HS Speaker and KODI streaming live TV on a Pipox7 (split to Windows 10 and Ubuntu server) using a multitouch wide screen openframe monitor. Never did install it as wall hole cutting stopped (except for the OmniTouch screens) and went to my Joggler Tabletops(15). Testing 6 of them in the family room and 5 in the master bedroom (ver low on the WAF)

That said here run Homeseer 3 on Ubuntu 16.04 64 bit and
Windows Server lite in an Oracle Virtual box on the same computer.

On said Virtual box I run 4 instances of speaker dot exe which connects to my audio stuff...other instances of HS3 on Linux also use the Virtual box speaker clients.

(as mentioned earlier been collecting SAPI speech fonts since the beginning of Homeseer...AT&T, Neospeech, (plus international speech to bug wife).