Using Tasker with VoxCommando

[laestrella]

Thank you so much for this. Just tested it out and it works flawlessly from my Moto 360 using AutoWear!!!!

I use Tasker, Autovoice and UDP Sender to send commands via Google Now to Voxcommando.

Tasker: https://play.google.com/store/apps/details?id=net.dinglisch.android.taskerm&hl=en
Autovoice: https://play.google.com/store/apps/details?id=com.joaomgcd.autovoice&hl=en
UDP Sender: https://play.google.com/store/apps/details?id=com.hastarin.android.udpsender&hl=en

Setup

Voxcommando: Ensure UDP listen port in Option > Advanced is set to 33000

Autovoice: Set Google Now Integration to Enabled. You may have to enable this first in your phone settings. (On Android Lolipop found at Settings > Accessibility > AutoVoice Google Now Integration set ON

Tasker: Go to Profiles and Click "+" to create a new profile. Then Select "Event", "Plugin", "Autovoice", "Recognized". Once there select Configuration and then Command Filter. Set your command filter to "%avcommsnofilter()", not including the quotations. Select OK and then click the Check Mark. Then select back to go to Event Edit.

You will then be asked to select an existing task or create a new one. We will select "New Task". You can name this task whatever you like. Then proceed by clicking the "+" the create a new task. From there select "Plugin", "UDP Sender" and then we will go to configuration. Set your destination IP to the machine where you have Voxcommando installed. Then set the port to "33000". Next we will go to the "Text" field and enter "vc.tellvox&& %avcommsnofilter()" no quotes. Also ensure you have a space between && and %.

You can now save that task and test it out. You can speak any command you have programmed into voxcommando. "Play" "Stop" "Play song whatever" "Play movie whatever".

Sometimes Google now will interpret me saying "Run KODI" as "Run Cody" so as a workaround I simply go to the command builder and add the word "Cody" to the phrases. I do this for any words that Google misinterprets.

Another thing you can do to get Google Now to work from any screen on your phone is go to Settings > Language and Input > Google Voice Typing > "OK Google Detection" and enable "From any Screen".

Hope this helps someone out there.

Cheers.

[RickyD333]

I use Tasker, Autovoice and UDP Sender to send commands via Google Now to Voxcommando.

Tasker: https://play.google.com/store/apps/details?id=net.dinglisch.android.taskerm&hl=en
Autovoice: https://play.google.com/store/apps/details?id=com.joaomgcd.autovoice&hl=en
UDP Sender: https://play.google.com/store/apps/details?id=com.hastarin.android.udpsender&hl=en

Setup

Voxcommando: Ensure UDP listen port in Option > Advanced is set to 33000

Autovoice: Set Google Now Integration to Enabled. You may have to enable this first in your phone settings. (On Android Lolipop found at Settings > Accessibility > AutoVoice Google Now Integration set ON

Tasker: Go to Profiles and Click "+" to create a new profile. Then Select "Event", "Plugin", "Autovoice", "Recognized". Once there select Configuration and then Command Filter. Set your command filter to "%avcommsnofilter()", not including the quotations. Select OK and then click the Check Mark. Then select back to go to Event Edit.

You will then be asked to select an existing task or create a new one. We will select "New Task". You can name this task whatever you like. Then proceed by clicking the "+" the create a new task. From there select "Plugin", "UDP Sender" and then we will go to configuration. Set your destination IP to the machine where you have Voxcommando installed. Then set the port to "33000". Next we will go to the "Text" field and enter "vc.tellvox&& %avcommsnofilter()" no quotes. Also ensure you have a space between && and %.

You can now save that task and test it out. You can speak any command you have programmed into voxcommando. "Play" "Stop" "Play song whatever" "Play movie whatever".

Sometimes Google now will interpret me saying "Run KODI" as "Run Cody" so as a workaround I simply go to the command builder and add the word "Cody" to the phrases. I do this for any words that Google misinterprets.

Another thing you can do to get Google Now to work from any screen on your phone is go to Settings > Language and Input > Google Voice Typing > "OK Google Detection" and enable "From any Screen".

Hope this helps someone out there.

Cheers.

Hey everyone, I am very new to VoxCommando. Just installed it yesterday. I did this whole process to the letter, but it's not completely working for me. I tap on my AutoVoice Recognize button on my phone to say my command, then I see VoxCommando on my desktop pc hear the command (it says "emulate: open gmail... open gmail is the command I want to execute), but it stops there. It doesn't open my gmail in my default browser, which is the one command that I have set up so far (except for the kodi plugin stuff I installed). I've tested it with VoxWav several times and it works. I'm not sure what I'm missing here.

[jitterjames]

Hi RickyD333,

Sounds like you are pretty close if VoxCommando is showing something in the history window.

Please enable logging in VoxCommando, then repeat the process with Autovoice and Tasker, and then upload your complete log as an attachment.

http://voxcommando.com/mediawiki/index.php?title=Log

As a side note, depending on your needs, you may find VoxWav an easier and better solution.

[RickyD333]

Thanks for the speedy reply jitterjames!

I got the attachment on here. If it didn't attach, let me know.

Like a lot of people I'm using VC to help automate my home. I started out using Tasker a while ago before I discovered VC. Had I learned about this sooner, I may have completely disregarded Tasker. This being the reason why I don't use VoxWav. However, I may end up making VC the primary tool/host for my home automation. The reason being that the microphone I wanted to install (Acoustic Magic Array I Microphone) throughout my house so I may be anywhere and give a command (I'm essentially trying to emulate the Star Trek computer/JARVIS), doesn't appear to support Android. Which is what I would need it to do if I am to use Tasker. I am assuming the communication between Tasker and VC can go both ways, so any function having to do with my Android phone will be activated through VC first (such as answering calls/receiving and sending text messages).

It's funny, I initially wanted VC so I can control my PC and XBMC/kodi. Looks like it may go further...

P.S. Are you the guy who did the VoxCommando Fireside Chat YouTube videos?

[jitterjames]

are you sure you are using:

Code: [Select]

vc.tellvox&& %avcommsnofilter()
I haven't tried to follow this tutorial myself but others seem to think it works...  I don't know for sure but I think %avcommsnofilter() should be giving the most likely phrase only.

In the VC log it shows that it is sending all possible phrases that Google could guess (separated by commas), not just the most likely one.

So it is trying to emulate the phrase "open gmail,hoping gmail,open gmail,hope in gmail,hopen gmail" which does not match the phrase "open gmail"

[jitterjames]

P.S. Are you the guy who did the VoxCommando Fireside Chat YouTube videos?

That's us! 

[nime5ter]

are you sure you are using:

Code: [Select]

vc.tellvox&& %avcommsnofilter()

We're not exactly the Tasker experts around here, but James, I think that %avcomm is the variable that only returns the first match. %avcomms returns a comma-separated list of matches. ... not sure what the "no filter" adds to the equation.

[RickyD333]

Yeah, it looks like I have that written out all correct. I'm attaching a screenshot of it.

And I loved those YouTube videos! I'm very much a visual learner, and the way those videos are done was perfect for me! You two are awesome!

[jitterjames]

Tasker: Go to Profiles and Click "+" to create a new profile. Then Select "Event", "Plugin", "Autovoice", "Recognized". Once there select Configuration and then Command Filter. Set your command filter to "%avcommsnofilter()", not including the quotations. Select OK and then click the Check Mark. Then select back to go to Event Edit.

You will then be asked to select an existing task or create a new one. We will select "New Task". You can name this task whatever you like. Then proceed by clicking the "+" the create a new task. From there select "Plugin", "UDP Sender" and then we will go to configuration. Set your destination IP to the machine where you have Voxcommando installed. Then set the port to "33000". Next we will go to the "Text" field and enter "vc.tellvox&& %avcommsnofilter()" no quotes. Also ensure you have a space between && and %.

I think these instructions are actually either incorrect or incomplete.  I think this is wrong:

Code: [Select]

vc.tellvox&& %avcommsnofilter()
It should probably be

Code: [Select]

vc.tellvox&& %avcommnofilter
Variables are explained here: http://joaoapps.com/variables/

AutoVoice
AutoVoice Recognized condition

%avcomms(): array with all the possible recognized voice commands
%avcommsnofilter(): array with all the possible recognized voice commands but without the command filter you set in the profile
%avcomm: first possible voice command that was recognized
%acommnofilter: first possible voice command but without the command filter you set in the profile
%avsource: Source of the voice command. Can be ‘normal’, ‘continuous’, ‘test’ or ‘googlenow’
%avword(): array with all the words of the first command that was recognized

[RickyD333]

Well I tried that and it still didn't work. In the little filter bar in the top of VC I get "emulate: ?ommnofilter"

[jitterjames]

I think you are doing something wrong in Tasker.  Without seeing it I can't tell what.

Can you please post a screen shot of the tasker setting?

[RickyD333]

I fixed it! It's so stupid too. 

In the AutoVoice section the Regex has to be selected.

[RickyD333]

Wait... what the heck?

It worked once... now it won't... I'm going to keep tinkering...

[RickyD333]

Here are some screenshots by the way.

[RickyD333]

Okay, so I keep trying things out and I got it to work again. However when I try to repeat it, it won't work. It seems like it is sort of random. I'm attaching two screenshots of my desktop that is running VC. One is labeled as using VoxWav, which works every time. The second image is labeled as using AutoVoice (and it is a shot of it matching the command, but not fulfilling it). I took these screenshots because that filter bar (is there a name for it?) at the top of the VC window looks different when it succeeds with VoxWav and when it does not complete the command with AutoVoice. I don't know if it means anything... but here it is.