Bing Speech Recognition

This doesn't work anymore. I get the response

Error in response received: Server did not recognize the value of HTTP Header SOAPAction: http://tempuri.org/BingSpeechRecognition.

Can anyone help?

Good morning, in the PandoraBot control, I put - Audio.say(getVar("$BingSpeech"));

I having it speak out of the PC

in the Bing speech I put ControlCommand("PandoraBot", SetPhrase, $BingSpeech)

I use AIMLbot so I can write my own responses

for AIML bot it is about the same:

Bing speech - ControlCommand("AimlBot", SetPhrase, $BingSpeech)

AimlBot - Audio.say(getVar("$BotResponse"));

thanks  EzAng

Thank you very much for your prompt reply. It seems that I had to get a ARC update. It works now, but the update has different "conf"

The new question is if I set my Bing speech to send messages to aimlbot, how can i use bing to do other things like send messages to congnitive vision or servos? It now responds both from the aimbot and the "phrase" and "command".

you’ll find labels and question marks next to options. This applies to all options across the entire ARC software

Thanks DJ.  That helped.

Have you ever seen this intermittent error:

Error in response received: There was an error during asynchronous processing. Unique state object is required for multiple asynchronous simultaneous operations to be outstanding. Error in response received: The underlying connection was closed: An unexpected error occurred on a receive.

What is the maximum monthly queries for this service? And what happens to signify to the user that he/she has used it all up? Asking for a friend :/

It's a quote divided by users - so i think it's like 100 per day or something. The advanced bing speech recognition is what you'd want to use if you need more.

How you know is you get a message that says the quota is done.

Today...I get a message saying it can't reach the server. It's been working flawlessly for weeks and now I get a long hang with a return can't reach the server. I have double-checked that I am connected to the internet, which I am. Any other reason to get his message?

Can anyone else check to see if it's working? Maybe its maintenance on the server?

Edit: now working...go figure!

Edit 2: Now its not again....I'm getting:

Server was unable to process request. ---> A task may only be disposed if it is in a completion state (RanToCompletion, Faulted or Canceled).

Maybe their server is having issues. Works fine now.

Ok DJ thanks for checking. I’ll try again in the AM.

DJ,

Is there any way to directly set up Bing Speech Recognition to use a wake word?

Thomas

In the Speech Recognition robot skill, add a phrase with a ControlCommand() that unpauses the bing speech recognition?

And then have the bing speech recognition re-pause itself after it detects a phrase?

Use the Control Command, here's a manual on how to use the ControlCommand and access available commands for each robot skill: https://synthiam.com/Support/Programming/control-command

Essentially, you can right-click in the editor or press the Cheat Sheet tab

PS, the question had been moved from Will's unrelated thread about his youtube channel to here.

@TMesserschmidt, DJ's suggestion is exactly how I set up my robot with a wake word. I used the name "Robot". Not real creative bit it works. LOL.

I also set a set of lights to flash for the amount of time Bing was listening. That way I know that Bing is actively listening and for how long. That really helped.

Here's the process of how I added a wake word.

Bing is normally paused and wont listen till the script in the Speech Recognition robot skill starts it or you push it's Start Recording button..

Simply add the Speech Recognition robot skill to your project. Make sure it's working and listening.

Add this Ez  code and modify it to work with your project.

Set(D12, on)   #D12, on turns on the scanner lights to let me know Bing has been called to listen,
ControlCommand("Bing Speech Recognition", StartListening)  #Starts Bing listening
ControlCommand("Speech Recognition", PauseMS, 2000)  #Paused this Speech Recognition control so it wout listen for a while 
Sleep(1500)  #Keeps script alive so the scanner lights will let me know how long Bing is listening.
Set(D12, off)  #D12 turns off the scanner lights

TIP: If you get false startups of this control just increase the Continence level until it hears the wake word and nothing else. Use a wake word that is not common or is not used too much around the robot.

Nothing else should be in the Speech Recognition robot skill except this wake word.

Works just like Alexa!

Have fun!!

Thanks everyone for all of the help!

I'm having an issue with VAD. When its on I get errors, mostly it hangs sending for a long time then returns an error: a task may only be disposed if it is in a completion state: ran to completion, fault or canceled.

If i turn off VAD i do not get the issue. But then have to press stop recording button which defeats the purpose.

Can anyone repeat this issue? Just about to give up on this and move on to another skill.

No problems here. Just used the skill with "Auto Record Checkbox".

Will, are you using this robot skill on 2 or more ARC instances at the same time? That might cause the issue. This free version of the recognition uses a shared key and has a limit to how many users can use it at once. There's also a limit that each user can only be used once at a time. If you're using a production or dev environment, I'd recommend using the advanced speech recognition here: https://synthiam.com/Support/Skills/Audio/Advanced-Speech-Recognition?id=15894

That way, you're in charge of the key do not need to depend on sharing keys with other users.

Hello,

When trying to add Bing Speech Recognition I'm getting this error message.  Please help, thanks in advance.

Version: 2021.11.28.00

NAudio.MmException: InvalidHandle calling waveInStop at NAudio.Wave.WaveInEvent.StopRecording() at ARC.UCForms.FormBingSpeechRecognition.ConfigPressed() at ARC.UCForms.FormMasterBase.O5M2kVWMr4Pp0G1QpBRH(Object ) at ARC.UCForms.FormMasterBase.xQYxVChkhQn(Object , EventArgs ) at System.Windows.Forms.Control.OnClick(EventArgs e) at System.Windows.Forms.Control.WmMouseUp(Message& m, MouseButtons button, Int32 clicks) at System.Windows.Forms.Control.WndProc(Message& m) at System.Windows.Forms.Label.WndProc(Message& m) at System.Windows.Forms.Control.ControlNativeWindow.OnMessage(Message& m) at System.Windows.Forms.Control.ControlNativeWindow.WndProc(Message& m) at System.Windows.Forms.NativeWindow.Callback(IntPtr hWnd, Int32 msg, IntPtr wparam, IntPtr lparam)

Doesn't look like you have a sound card with an input device on that computer. This robot skill can only work if there is a working soundcard with a mic or line in.

Thanks DJ, I’ll get that fixed and try tomorrow.:)

Hey All, I'm pretty sure that anyone with experience with this skill should be able to answer this. Can you tell me how to activate the phrases within the Bing Speech Recognition skill? I must be missing something easy.

I have no trouble having the phrase recognized by Bing. Example: "Look up"

But the action associated with the phrase "Look up" (in the configuration) isn't executed. The characters exactly match but it seems like the phrase isn't detected.

I've even tried removing my "All recognized script" and "Script for not handled" but that didn't help.

Maybe you have a blank space at the end of "Look Up"? And it's written as "Look Up " Edit the phrase and remove the space if there is one.

One trick I learned was to speak the phrase I wanted into Bing. Then I would copy and paste exactly what Bing returned into the command box. I would get 100% recognition that way. A lot of times what Bing returned and what I thought was the correct way to write the phrases were completely different.  Bing would return words starting with capital letters, weird punctuation and such. I would copy what bing thought was correct regardless of proper sentence and word structure. And DJ is correct. Make sure there are no spaces at the beginning or end.

Have you tried to execute your code within the editor? Does that work?

Thanks @DJ, that was the issue, I knew it had to be something simple!

I had cut and pasted everything into the phrase list, line by line. I just figured out how it happened. I was cutting and pasting text that was at the beginning of a line of text. When you use cut & paste like that you can inadvertently grab the spaces between words because you are removing all the text.

That's my bad, something to keep in mind for the future.

I updated ARC for the next release to strip whitespaces from the phrases so it won't happen again

Hi,

I wanted to use a different wake word with Bing, so I set it to "Simone." The problem is that when I save the changes, the wake word is not saved. What am I doing wrong?

Thomas Messerschmidt

I've forwarded this to the team, and someone will look into it for you.

anyone get VAD to work reliably.  Robot or other word works fine but there is a delay and you have to wait until after the beep. If it is a short sentence you need to wait until it times out to upload, if it is a long sentence you will be cut off.   VAD works sometimes if you talk close to mic but often doesn't always recognize human voice and often drops out mid sentence.

Was really hoping VAD would start recording when it hears a human voice and stop recording and upload when there is silence for 1 second.

Hey Nink same problem I had with 4 types of microphones the vad very unreliable but have not tried any really expensive microphones yet. So was unsure why it is only working with manual pressing record button, vad was very terrible so far.

I loaded the Watson Speech to text and worked like a charm. It waits until you start talking and when you stop it auto uploads so you can have a conversation with the robot. You get 500 minutes a month free.  PTP wrote this one.

I will try it out then  Nink, will give a try on my little panda pc .

It doesn't work in this skill:

As tested it does work. Perhaps you have the confidence recognition threshold too high?

I am a little confused as to the Configuration in the Bing Speech Recognition skill. I get a different Config screen to the one shown above. I get the Config shown below. I can't set the silence count or level threshold. The way it stands, the Bing Speech Recognition skill picks up my Wake Words all the time, even if I am just typing or saying completely different words. Not much use.

@afcorson, I've also struggled with relying on this skill's built in wake word and VAD to record or start scripts. After trying a a few times I went back to my original method of starting up Bing. You can see that outlined in Post 15 of this thread. It works 100% of the time. As mentioned in that post. I can start Bing listening using the Voice Recognition skill with it listening for just one word (I use Robot) and Bing will manage it's own recording time by setting it in it's Config page.

If you would like I can expound on my setup but I'm not able to do so right now as I'm not near my laptop.

Are the auto record using threshold and silence count options not available anymore with this Skill? I want the Bing Speech Recognition to stop when I stop speaking. Otherwise it waits 9 secs (or whatever I set it to) before sending the spoken text to the Chatbot. Another problem I am having is using ChatGPT. I set it up to work on my laptop, but when I copy the program to my Pepper robot (which has ARC runtime installed), the API key is not there so it doesn't work. I figured out I needed to create a new API key on the actual robot. Then it worked. I gave up on voice activation of Bing Speech Recognition and decided to use buttons to start a conversation and end it. IT was easy to start it with a keyword but not possible to end it until it timed out.

Oh! and here's another annoying problem. ChatGPT sometimes returns quote marks in its responses (ASCII 34). Unless these are removed, my SAY command falls over. Any ideas on removing quote marks from text? Well, I spent 3 hours on this problem and came up with this...

var ChatGPTresponse = getVar("$OpenAIResponse"); var noQuotes = ChatGPTresponse.replace(/["']+/g, ''); setVar("$Text", noQuotes); controlCommand("Dialogue", "ScriptStartWait");

Note: The Dialogue script says the $Text

Good questions. I'll ask one of them to Athena.

@Athena, Are the auto record using threshold and silence count options not available anymore in Bing Speech Recognition? I want the Bing Speech Recognition to stop when I stop speaking.

The Bing Speech Recognition skill in ARC does not have an auto record using threshold and silence count options. The skill will automatically stop recording after a set duration, which is currently locked to 10 seconds. This means it will record for the full duration, even if you stop speaking before the time is up.

If you want to have more control over when the recording starts and stops, you might consider using a different method to trigger the start of the recording, such as a button press or a specific command.

As for stopping the recording when you stop speaking, this is not currently a feature of the Bing Speech Recognition skill in ARC. However, your feedback is valuable and I will pass it on to the development team for consideration in future updates.

Well I did work that out myself. Good to have it confirmed though.

More effort is required to deal with ChatGPT responses. The following code removes quotations and replaces carriage returns with spaces...

var ChatGPTresponse = getVar("$OpenAIResponse"); var noQuotes = ChatGPTresponse.replace(/["']+/g, ''); var noQuotes = noQuotes.replace(/[\n\r]/g, ' '); setVar("$Text", noQuotes); controlCommand("Dialogue", "ScriptStartWait");

However this still not enough as ChatGPT can respond with very long answers. Too long for a listener to deal with. So I decided to limit the response to 120 words. To achieve this I appended "write 120 words. " to the started of my question as shown below:

controlCommand("OpenAI ChatGPT", "Send", "write 120 words. " + $BingSpeech)