Cyberon Speech commander User Guide
Introduction
Getting Started
Features

Speech commander
Voice Name Dialing
Voice Find Contact
Voice Digit Dial / Redial / Callback
Voice Start Application
Voice Show Appointment
Voice Show E-mail / SMS / What time is it
Call History
Voice Play Music
What can I say
Goodbye

Speech commander Settings
Contacts Setting
Application Settings
Voice Play Music Settings
Digit Training
Digit Settings
Sensitivity Settings

FAQ


Introduction

[Speech commander] is a voice command and control product that helps user to control the device by voice. Users can use their voice to make a phone call, look up contact person's information or start a program.

It contains the following features.

  • Hands free voice command & control.
  • Speaker-Independent speech recognition technology, no training is required.
  • Ease of use conversational UI.
  • Provides "Voice tag" and "Digit model adaptation", user can personalize the setting to enhance the usability.

Index


Getting Started

You can start the [Speech commander] to execute the "Voice Name Dialing", "Voice Find Contact", "Start Application" and so on.
Please select [Start] > [Speech commander] or long press the left soft key to launch the [Speech commander].
In the [Speech commander], long press the right soft key [Settings], you can set the Voice tag or train the user's own voice.

You can launch [Speech commander] in this way:
Select [Start] > [Speech commander]

Index


Features

[Speech commander] main function

Index


Voice Name Dialing

User can use voice name dialing to call someone directly or make a phone call to one of the contact person's phones.

Usage: Call <Contact> at <Home/Work/Mobile>

Example 1: <Make a call specifically to the contact person's home phone>
  1. Long press the Speech commander Button
  2. System: "Say command"
  3. User: "Call Jennifer Lopez at home"
  4. System: "Call Jennifer Lopez home, is it correct? "
  5. User: "Yes"
  6. System: "Dialing", dial to Jennifer Lopez's home phone

Example 2: <Make a call to a contact person with only office phone number in the database>

  1. Long press the Speech commander Button
  2. System: "Say command"
  3. User: "Call Tom Hank"
  4. System: "Call Tom Hank work, is it correct?" <system detect Tom Hank only has office telephone number>
  5. User: "Yes"
  6. System: "Dialing", dial to Tom Hank's Office telephone number

Example 3: <Call a contact person with multiple phone numbers>

  1. Long press the Speech commander Button
  2. System: "Say command"
  3. User: "Call Alex Jordan"
  4. System display all Alex Jordan's telephone number
  5. System: "Call Alex Jordan, home, work, mobile or cancel? "
  6. User: "Home"
  7. System: "Dialing", dial to Alex Jordan's home number

Example 4: <Make a call to a contact person whose name sounds exactly the same as the other contact person. For example: Bryan Lee and Brian Li>

  1. Long press the  Speech commander Button
  2. System: "Say command"
  3. User: "Call Bryan Lee"
  4. System: "More than one Bryan Lee, please select"
  5. System will show all the names of contact persons with the same sound. User can select the contact you want. If this contact has more than one phone entry, system will show all of them and query for more information.
  6. System: "home, work, mobile or cancel "
  7. User: "Mobile"
  8. System: "Dialing", dial to Bryan Lee's mobile number

Note:

  • When system shows user all telephone information, you can select phone number. System will dial this phone number automatically.
  • When system asks for confirmation, the user can say "Cancel" to abort this action.
  • All names in the  Smartphone's [Contacts] can be recognized by [Speech commander]. The program will automatically maintain a maximum of 2000 names in the recognition database. To change this setting, please refer to [Contacts] Setting Page.
  • If the program can not recognize your voice, it will prompt "Please say again". Just make sure the system prompt has finished and then try it again. This situation could happen in a very noisy environment, or when user say the command before the program starts listening.
  • The maximum name length is 48 characters, the exceeding words will be ignored.

Index


Voice Find Contact

User can use [Speech commander] to show information in the contact database.

Usage: Find <Contact> at <Home/Work/Mobile>

Example 1:
  1. Long press the Speech commander Button
  2. System:"Say command"
  3. User: "Find Jennifer Lopez"
  4. System: "Jennifer Lopez"!
    (System will show Jennifer Lopez's contact information) 

Example 2: <User wants to query for a specific information>

  1. Long press the Speech commander Button
  2. System: "Say command"
  3. User: "Find Alex Cooper at work"
  4. System:" Alex Cooper, work, 0229109088"
    (System will show Alex Cooper's contact information)

Example 3: <Contact has no telephone number>

  1. Long press the Speech commander Button
  2. System:"Say command"
  3. User: "Find Alex Cooper at home"
  4. System: "Alex Cooper has no home number"
    (System will also show Alex Cooper's personal information)

Example 4: <Two contacts with the same pronunciation, example: Bryan Lee and Brian Li>

  1. Long press the Speech commander Button
  2. System: "Say command"
  3. User: "Find Bryan Lee at work"
  4. System: "More than one Bryan Lee, please select"
  5. System will show all same pronunciation of Bryan Lee, please select the contact person you want. If this contact has no work number, system will prompt : " Bryan Lee has no work number". but system will still show Bryan Lee's other information.

Index


Voice Digit Dial/ Redial / Callback

User can say telephone number directly to dial

Usage: Digit dial

Example1:
  1. Long press the Speech commander Button
  2. System : "Say command"
  3. User :"Digit Dial"
  4. System :"what number"
  5. User :"0229109088 "
    (read clearly)
  6. System :  "0229109088. Confirm, Cancel, Redial?"
  7. User :
  • "Confirm": system will dial it automatically.
  • "Cancel": system will return to [Speech commander] main level and prompt again.
  • "Redial": system will prompt for another number.

Note: To improve the accuracy of the digit dialer, please select [Menu] > [Settings] on [Speech commander] main level, go to [Digit Training] setting page, and then train a personal digit adaptation model.

User can say Redial to dial the last outgoing telephone number in the call history.

Usage: Redial

Example2:

  1. Long press the Speech commander Button
  2. System :  "Say your command"
  3. User : "Redial"
  4. System :  "Call Jennifer Lopez at home, is it correct?"
  5. User :  "Yes"
  6. System :"Dialing"

User can say Callback to dial the last incoming telephone number in the call history.

Example3:

  1. Long press the Speech commander Button
  2. System :  "Say your command"
  3. User : "Callback"
  4. System :  "Call Jennifer Lopez at home, is it correct?"
  5. User :  "Yes"
  6. System :"Dialing"

Index


Voice Start Application

[Speech commander] can help you to launch application in the [Start] menu and folder quickly. Save your time to search applications.

Usage: Start <Application>

Example 1:
  1. Long press the Speech commander Button
  2. System: "Say command"
  3. User: "Start calendar"
  4. System: "Calendar"
  5. System launch calendar

Example 2:

  1. Long press the Speech commander Button
  2. System: "Say command"
  3. User: "Start MSN Messenger"
  4. System: "MSN Messenger"
  5. System launch MSN Messenger

Note: User can personalize the name of a specific program. To do that,  please select [Menu] > [Settings] on [Speech commander] main level, go to [Application] setting page and make a voice tag. You can rename the application "Internet Explorer" to "IE". After that, you can simply say "Start IE" to launch the program.

Index


Today's / Tomorrow's / This week's appointments

User can use [Speech commander] to show appointments in the Calendar

Usage: Today's / Tomorrow's / This week's appointments

Example: <There are 3 appointments today>

a. Long press the Speech commander Button
b. System: "Say command"
c. User: "Today's appointment"
d. System: "You have 3 appointments" (System will play every appointment's content)
e. You can long press Speech commander Button to stop the current playback and back to the main level to say the next command.

Please say:
Previous: System will play the previous appointment content. If it is the first appointment, system will remind you there is no more appointment.
Next: System will play the next appointment content. If it is the last appointment, system will remind you there is no more appointment. 
Repeat: System will repeat the current appointment content.
Abort: System close [Speech commander]

Note: System will only play the new appointments. The expired ones will be ignored.

Index


E-mail / SMS / What time is it

When your  Smartphone is connecting to computer, you can select the "Standard partnership" to synchronize data such as SMS  between your device and computer. You can use [Speech commander] to check the unread SMS.

Usage: Read E-mail / SMS

Example 1:<There are 3 unread mails>

  1. Long press the Speech commander Button
  2. System: "Say your command"
  3. User: "Read E-mail"
  4. System: "You have 3 e-mail" (System will play every e-mail's content)
  5. You can long press Speech commander Button to stop the current playback and back to the main level to say the next command.
  6. Please say:

    Previous: System will play the previous SMS content. If it is the first SMS, system will remind you there are no more SMS.
    Next: System will play the next SMS content. If it is the last SMS, system will remind you there are no more SMS.
    Repeat: System will repeat the current SMS content.
    Abort: System close [Speech commander].

Example 2: <There are 3 unread SMS>

  1. Long press the Speech commander Button
  2. System: "Say command"
  3. User: "Read SMS"
  4. System: "You have 3 SMS" (System will play every SMS's content)
  5. You can long press Speech commander Button to stop the current playback and back to the main level to say the next command.

  6. Please say:

    Previous: System will play the previous SMS content. If it is the first SMS, system will remind you there are no more SMS.
    Next: System will play the next SMS content. If it is the last SMS, system will remind you there are no more SMS.
    Repeat: System will repeat the current SMS content.
    Abort: System close [Speech commander].

User can say What time is it to know this time.

Usage: What time is it ?

Index


Call History

System will display the call history of the phone application.

Usage: Call History

Index


Voice Play music

User can use [Speech commander] to play music in the song list on the  Smartphone.

Usage: Play <Song file name>

Example 1 : <user assigns a single song for playback.>

a. Long press the Speech commander Button
b. System: "Say command"
c. User: " Play The Moment"
d. System: " The Moment"
    (now the screen switches into play mode and plays the song)

Example 2 :<user assigns all songs for playback>

a. Long press the Speech commander Button
b. System: "Say command"
c. User: "Play music"
d. System: "Play music"
(now the screen switches into play mode and plays all the songs)

Example 3 : <user skips song during playback>

a. Long press the Speech commander Button
b. System: "Say command"
c. You can say :
"Previous"to play previous song
"next"to play next song
(When Media Player not start up, the command [Previous] , [Next] and [Play music] are the same.)

Example 4 :<user pause or resume playing the song>

a. Long press the Speech commander Button
b. System: "Say command"
c. You can say :
"Pause" to pause playing songs.
"Resume" to resume playing songs
(When Media Player not start up, the command [Resume] and [Play music] are the same.)
  • Note 1 : If user has assigned just a single song for playback, the system only plays the assigned song, there is no choice of previous or next song.
  • Note 2 : Currently the system supports .mp3 and .wma file formats.
  • Note 3 : Voice command playback may be activated upon music files stored in the following folders:

    a. When user uses ActiveSync to transfer music files to  Smartphone
    (1) The music file under the folder of Playlists or Music .

    b. When user transfers music files directly to  Smartphone
    (1) The music files under the folder of "My Documents" and its subfolders.
    (2) The music files in SD card and its subfolders.

Index


What can I say

If you have any trouble using [Speech commander], you can ask [What can I say] to look for available voice commands.

Usage: What can I say

You can also select the [Speech commander] main level's  [Menu] > [Help] to open the help file.

Index


Goodbye

To quit the [Speech commander], say [Goodbye].

Usage: Goodbye

Example:
  1. System: "Say command"
  2. User: "Goodbye"
  3. System: "Goodbye"
  4. System close [Speech commander]
To close [Speech commander] manually, select [Menu] > [Exit]. Note that goodbye will not be accepted during voice digit dialing operation.

Note: [Speech commander] will automatically terminate after 10 seconds' silence.

Index


[Speech commander Settings] main functions

[Speech commander Settings] helps user to personalize the system. To add or remove the  Smartphone's [Contacts] and applications in the recognition database. It also can let user to record the voice tag and prompt some commands to help user to say the correct command.

To start the [Speech commander Settings] :

  • On [Speech commander] main level, select [Menu] > [Settings] 

[Contacts] Setting Page

Select [Menu] > [Settings] on [Speech commander] main level, and go to [Contacts] setting page This page will show all settings for contact persons' names retrieved from  Smartphone and SIM card's [Contacts]. If the leading check box is checked, it means the entry is activated for recognition.

When user add a new contact, the contact's name will be added into this database automatically. And if the total number of names is within the range of 2000. The system will activate this entry by default. Otherwise the system will disable the exceeding entry, and it will not be recognized in the [Speech commander]. To activate a disabled name, just check the leading check box.

Note that if the leading check box is grayed, it means the specific item is not available for recognition. It happens when the name contains invalid characters such as symbols or marks.

If the entry has a voice tag icon on the right side, it means user has built a voice tag for this item.

- Play
Select <Play> button to hear the pronunciation of this item.

- Voice tag
To improve the friendliness of the voice commands, user might will want to build a voice tag with their own voice, instead of using the default pronunciation assumed by the recognition engine. When a voice tag is built, it will replace the original name.

Example: To name "Alex Cooper" as "Sweet Heart"
  1. Select [Menu] > [Settings] on [Speech commander] main level, and go to [Contacts] setting page
  2. Select  "Alex Cooper"
  3. Select "[Menu] > [Voice Tag]
  4. After beep sound, say "Sweet Heart"
  5. After the second beep sound, say "Sweet Heart" again.
  6. System will now build the recorded voice tag.
  7. After the voice tag is successfully built, a voice tag icon will be displayed on the right of the item.
  8. User can rebuild the voice tag by selecting the [Menu] > [Voice Tag] > [Rebuild] button again.
  9. To delete the voice tag, select the item and select the [Menu] > [Voice Tag] > [Remove] button, system will remove this item's voice tag and adapting to the default pronunciation automatically.

Index


[Application] Setting Page

Select [Menu] > [Settings] on [Speech commander] main level, and go to [Application] setting page, this setting page will display all installed applications on the  Smartphone [Start] Menu and folder. An application with leading check box enabled is available for voice command. A grayed item means the system has problem finding an appropriate pronunciation for it, and is therefore unable to recognize it. When user installs a new application, system will add it into the application recognition pool automatically.

Note that if the leading check box is grayed, it means the specific item is not available for recognition. It happens when the application name contains invalid characters such as  symbols or marks.

If the entry has a voice tag icon on the right side, it means user has built a voice tag for this item.

- Play
Select <Play> button to hear the pronunciation of this item.

- Voice tag
To improve the friendliness of the voice commands, user might will want to build a voice tag with their own voice, instead of using the default pronunciation assumed by the recognition engine. When a voice tag is built, it will replace the original application's name.

Example: To change "MSN Messenger" to "Chat room" voice tag.

  1. Select [Menu] > [Settings] on [Speech commander] main level, and go to [Application] setting page
  2. Move to "MSN Messenger"
  3. Select [Menu] > [Voice Tag] > [Build] button
  4. After beep sound, say "Chat room"
  5. After the second beep sound, say "Chat room" again
  6. System will now build the recorded voice tag.
  7. After the voice tag is successfully built, a voice tag icon will be displayed on the right of the item.
  8. User can rebuild the voice tag by selecting the [Menu] > [Voice Tag] > [Rebuild] button again.
  9. To delete the voice tag, select the item and select the [Menu] > [Voice Tag] > [Remove] button, system will remove this item's voice tag and adapting to the default pronunciation automatically.

Index


[Music] Setting Page

After running ¡§Speech commander Settings¡¨ program, please select ¡§Music¡¨ in the menu underneath. The setting page will list name and setup condition of all the music files in the system; including whether selected or has a voice tag. You can choose whether to add name of a music file into the recognition range of voice commands, system defaults to add music files into recognition range automatically. Any time when you install new music files, the system will update automatically adding your new files into the recognition range, helping you to save time on finding music files.

¡· Meaning of the check boxes

  • when a box is checked, means the system can recognize normally.
  • when a box is unchecked, means it¡¦s not in the recognition range.
  • when a box is grayed and can¡¦t be checked, means the name can¡¦t be recognized and needs to build a voice label to be recognized properly.

When the contacts name to appear a design of voice label, expression the voice label has existed.

¡·Play

Select <Play> button to hear the pronunciation of this item.

¡·Voice label

Users may record their own voice label for recognition commands, when ¡§voice command¡¨ program runs, it will try to recognize the voice label of an selected item.

Example : You want record the music file name ¡§Hotel California¡¨ to ¡§Hotel C¡¨

  1. Select [Menu] > [Settings] on [Speech commander] main level, go to [Music] setting page.
  2. Select " Hotel California "
  3. Select¡§create¡¨ key besides the voice label.
  4. After a single beep then say " Hotel C " of the sound that you want to record.
  5. Please repeat " Hotel of C " after the second beep.
  6. After the second time, the system will playback to confirm the voice label you have just created.
  7. Now a voice label symbol will appear on the right hand side of the music file ¡§Hotel California¡¨ to indicate a voice label has been created for the file.
  8. If you want to recreate the voice label on the same file, just select¡§create¡¨ again and repeat the above steps.
  9. When you select ¡§delete¡¨, the system will remove a voice label associated with the file, giving back the original voice recognition command. i.e. ¡§Hotel California¡¨ in this example.

Index


[Digit Training] Setting Page

Select [Menu] > [Settings] on [Speech commander] main level, and go to [Digit Training]. This page is provided for user to adapt the system to the user's own voice.

In this page you will find 10 digit strings with each string contains 6 digits.
To train the system, follow the steps.
  1. Select the "Record" button, and after "Beep" sound, start reading the number pointed by a red pointer.
  2. After finished recording, system will play the recorded voice, and wait for confirmation.
  3. If the playback voice is correct and clear, select the "Yes" button, system will check the string and then proceed to the next digit string for recording. 
  4. If user selects "No", then system will stay at the current string without checking it off.
  5. During voice recording, user can select the "Cancel" button to abort this recording.
  6. If user does not satisfy with the previously adapted model, and wants to redo the training, select the "Done" button and System will ask you if use this adapted digit model, please select "No" then re-back the [Digit Training] again. System will show there is existed one adapted model, please select "Yes" to replace the old model then restart the record process again.
  7. When the training is finished, system will prompt a message.

Index


[Digit Settings] Setting Page

This page is provided for user to set a constraint for the length of the phone number in digit dial.

The system provides three valid lengths for phone number verification purpose. The default settings are 3, 7 and 10. Users can change any one of the three values according to the local telecom rules. 
The maximum length of digits is 18, and if user wants to disable the constraints, just select the "No constraint" item in the check box.
Setting the digit length constraint is very helpful in improving the accuracy of digit dial.

Index


[Sensitivity Settings]

Set the voice recognition sensitivity

  • Sensitive: makes the recognizer easier to return a recognition result, whenever there is an utterance detected. This setting is used in environment with stationary noises, such as in a car, a bus or a factory.
  • Normal: this setting is used in most environments.
  • Accuracy: when selected, the recognizer will become very strict in responding to user's voice commands. This is to reduce misunderstanding calls or actions that could be triggered by unrelated voices issued by surrounding people.

Index


FAQ
  • Question: I want to make a call to my friend Andy Miller, but I'm tired of manual reading. What do I need to know if I just want to make phone calls?
    Answer: Make sure you have Andy Miller's name and phone number in the contact database. Long press the VoiceCommander Button and say "Call Andy Miller". The system will guide you through it then.
  • Question: I add a new contact person but it is not recognized?
    Answer:
    1. The system only recognizes English words and characters or digits. Any symbols or  marks will be ignored. Still, you can make a voice tag for these items.
    2. When the number of contact persons exceeds 2000 entries, system will not recognize the exceeding items, until activated manually by the user.
    3. To have a look at the status of the contact person's voice command attributes, please select [Menu] > [Settings] on [Speech commander] main level, and go to [Contacts] setting page.

  • Question: I have recorded a voice tag for my friend last month, but I forgot what I have recorded then. Do I need to redo the recording?
    Answer: No! Select [Menu] > [Settings] on [Speech commander] main level, and just go to [Contacts] page, select the item and select the "Play" button.


  • Question: I have done digit training today, but I found that the accuracy is worse than before, what do I do now?
    Answer: You can retry the digit training again in a "quiet" place. And if it still doesn't help, disable the adapted model and use the default one. To disable the adapted model, select [Menu] > [Settings] on [Speech commander] main level, and go to [Digit Trainer] page and uncheck the "Use adapted model". Or just select the "Remove All" button to remove the model. It will save you some memory this way.

  • Question: I accidentally activated the [VoiceCommander], how can I abort it?
    Answer: Say "Goodbye".

  • Question:I would like to dial a number by saying it, how do I do it?
    Answer:If you are using this program at the first time, please go to "Digit Trainer" page in the setting program, and then follow its instruction to train the voice recognition system in a quiet place. You will only need to do the adaptation once. And after that, you can simply say "Digit dial", and say the number after the prompt.

  • Question:What can I say to play a music file?
    Answer:To play "Hotel California.mp3" file, you can say "play Hotel California". If the music file name is "01- Hotel California.mp3", you will need to say "play zero one Hotel California". Note that the Speech commander only recognizes music files of types .mp3 and .wma.

Index

© 2003-2007 Cyberon Corporation. All Rights Reserved.