Bruno Speech Recognition BETA v1.2

FunkDaMonkMan · 04-22-2005, 11:55 PM

Bruno Speech Recognition 1.2

This program will accept speech commands and output a variety of system commands.

Prerequisites - Microsoft Speech Recognition Software (comes with MS office)
Read the "Read Me.txt" to find out how to install this.

The program can do the following:

-type any key on the keyboard, and hold modifier keys (shift, ctrl, alt)
-open any program
-close any specific program, or all programs matching a string (ex. "internet explorer")
-pause for any given amount of time
-put multiple operations together into one command (ex. Open internet explorer, wait 4 seconds, type hello, pess enter, wait 2 seconds, close internet explorer)

The grammar is configured by an INI file which can be handwritten, or you can use the program's automated grammar builder.

There are two levels of commands. Master and child commands. It is possible to have a master command of "open" and chlid commands of "internet explorer" and "notepad". If the user says "open notepad" or "open internet explorer" those programs will open.. but if the user just says "notepad" nothing will happen.

It is also possible to make commands with just one word, such as "internet".

In addition to all of this, it is possible to load 5 grammar files at once, select which ones are active, and even control the program itself with voice commands. (ex. "Enable Speech" "Disable Speech" "Disable Five" "Enable Three")

This program is beta, and you may find errors. I have only tested it on XP Please notify me VIA email or post here if you find any bugs. Also, please be aware that i quickly typed the Read Me, and I may have missed something.

Gobby · 04-23-2005, 04:17 AM

Not being rude or anything but how does this differ from NaviVoice?

FunkDaMonkMan · 04-23-2005, 10:43 AM

I have never used navivoice, sorry.

I will take a look at it.

FunkDaMonkMan · 04-23-2005, 10:49 AM

Ok the difference....

Navi-voice can output very simple text commands ex. "hello", and it is oriented toward navigation. Also, the commands must be manually entered into an xml file.

BSR can use tree's of commands that can be build at run-time. BSR supports complex commands, and commands that string together. My program isn't oriented toward any particular program, although I have found it useful for online gaming.

Navi-voice has many other features besides voice commands and is a very nice program, but the two programs do different things.

Bugbyte · 04-23-2005, 11:52 AM

This looks cool. I've had a lot of trouble configuring NaviVoice and have been looking for a way to control all of my apps by voice.

However, when I clicked on the link I got a screenful of garbage. Is there a particular browser required?

xBrady · 04-23-2005, 12:20 PM

Quote: Originally Posted by Bugbyte

This looks cool. I've had a lot of trouble configuring NaviVoice and have been looking for a way to control all of my apps by voice.

However, when I clicked on the link I got a screenful of garbage. Is there a particular browser required?

I got it to work in IE and Firefox.

Bugbyte · 04-23-2005, 01:23 PM

I was using Safari. I tried Firefox and downloaded the file. It's got a .msi extension on it. What's that?

FunkDaMonkMan · 04-23-2005, 01:27 PM

I'm working on adding a spelling feature...

ex. (Search for G-I-R-L-S), could open google and search for girls. Again, this is configurable, so it could also open winamp/frodo and search for song title, if you set it up this way.

Would anyone find this feature useful?

FunkDaMonkMan · 04-23-2005, 01:28 PM

Quote: Originally Posted by Bugbyte

I was using Safari. I tried Firefox and downloaded the file. It's got a .msi extension on it. What's that?

Its an install file.

0l33l · 04-23-2005, 01:29 PM

Quote: Originally Posted by FunkDaMonkMan

Ok the difference....

Navi-voice can output very simple text commands ex. "hello", and it is oriented toward navigation. Also, the commands must be manually entered into an xml file.

That's why I have macros. They do the exact same things

I'd love to know how you add commands to SAPI at runtime

FunkDaMonkMan · 04-23-2005, 02:10 PM

Quote: Originally Posted by 0l33l

I'd love to know how you add commands to SAPI at runtime

YGPM

xBrady · 04-23-2005, 03:28 PM

Quote: Originally Posted by Bugbyte

I was using Safari. I tried Firefox and downloaded the file. It's got a .msi extension on it. What's that?

It's an install file used by Windows installer.

Bugbyte · 04-26-2005, 12:56 PM

Downloaded your app and started using it today. It works well and the best part was that I was able to easily configure it in about 10 minutes! I figured out the "parent-child" pretty quickly.

I let it running when I shut the car off so we'll see if it survives hibernation. That's been the biggest problem I've had with apps like Centrafuse.

Stupid, but exciting - I was able to use voice to automatically open and close my screen while driving! I bet the folks in rush hour were wondering what the hell was going on. Unfold. Fold. Unfold. Fold.

One suggestion is to code in an overarching "keyword" for the app to use so it ignores you unless you use that word "Computer, open screen". I coded it in as a parent line but that means for each grammar file I have to put it in.

An additional suggestion is to give it a user selectable option to respond to you "Computer open screen" "Executing command". Also, some feedback at the bottom of the window to let you know that it heard you even if it didn't understand you. It does tell you when it understands you but otherwise you can't tell if the app is working or not. Something like: "Unable to decode" or maybe even let the user put the text in. This should also be able to be spoken.

Also, I know it can send key strokes but is there a way to sendkeys to an application? I use FrodoSatX and it doesn't have hotkeys but you can use sendkeys. This might be useful for other apps.

Thanks for the beta. I've found it easy to use and so far no bugs, but we'll see today when I start the car and say "iBug, open screen"

xBrady · 04-26-2005, 01:01 PM

Quote: Originally Posted by Bugbyte

Stupid, but exciting - I was able to use voice to automatically open and close my screen while driving! I bet the folks in rush hour were wondering what the hell was going on. Unfold. Fold. Unfold. Fold.

OMG that is cool!

FunkDaMonkMan · 04-26-2005, 07:58 PM

Quote: Originally Posted by Bugbyte

Downloaded your app and started using it today. It works well and the best part was that I was able to easily configure it in about 10 minutes! I figured out the "parent-child" pretty quickly.

I let it running when I shut the car off so we'll see if it survives hibernation. That's been the biggest problem I've had with apps like Centrafuse.

Stupid, but exciting - I was able to use voice to automatically open and close my screen while driving! I bet the folks in rush hour were wondering what the hell was going on. Unfold. Fold. Unfold. Fold.

One suggestion is to code in an overarching "keyword" for the app to use so it ignores you unless you use that word "Computer, open screen". I coded it in as a parent line but that means for each grammar file I have to put it in.

An additional suggestion is to give it a user selectable option to respond to you "Computer open screen" "Executing command". Also, some feedback at the bottom of the window to let you know that it heard you even if it didn't understand you. It does tell you when it understands you but otherwise you can't tell if the app is working or not. Something like: "Unable to decode" or maybe even let the user put the text in. This should also be able to be spoken.

Also, I know it can send key strokes but is there a way to sendkeys to an application? I use FrodoSatX and it doesn't have hotkeys but you can use sendkeys. This might be useful for other apps.

Thanks for the beta. I've found it easy to use and so far no bugs, but we'll see today when I start the car and say "iBug, open screen"

I can easily impliment talkback. Also, I will put in a "hypothesis" bar that will display what it thinks you are saying, even if it doesn't confirm it.

I am almost finished with a spelling feature, that will let you type hands free.
This will all be on the next release.

btw, do you think that the response should be "Command recognized" or.. just simply say what the command is. "Open screen"

also.. i'm not sure if i documented this, but if you make the master command "independent", then you don't have to say the master command, just the child command.