Release Notes for Dragon Medical SpeechKit 3.0 Service Release 16 (Android edition)

Dragon Medical SpeechKit is a component of the Nuance Healthcare Development Platform. Rebranding the components is currently an ongoing activity. Dragon Medical SpeechKit was previously called 360 | SpeechAnywhere Services; please be aware that references to 360 | SpeechAnywhere Services 3.0 apply to this release.

New in service release 16

• Dragon Medical-specific swipe gestures, to control recording (swipe-to-speak) and remove text (swipe-to-delete), have been removed. Standard UX guidelines for mobile devices designate the swipe gesture for sliding menus; Dragon Medical SpeechKit no longer interferes with this functionality in your application. [DNB-7916]

Solved issues in service release 16

• If a session is not closed explicitly, the Android edition no longer causes unnecessary network traffic when no recording is running. [DNB-7836]

Important information

• Dragon Medical SpeechKit (Android edition) is compatible with server versions 2.0, 2.1 and 3.0.
• Dragon Medical Server 3.0 is compatible with client versions 2.1 and 3.0.
• Information on the voice commands changed with each version of Dragon Medical Server is now available in the end user help.
• On the speech bar the colors for recording on/off have changed from red (on) - gray (off) to green (on) - red (off).
• If your application targets Android 5.1 or lower (targetSdkVersion 22 or lower), a dependency must be added to the version 4 Support library revision 23 or higher. [DNB-5740]

New in version 3.0

Manage Vocabulary (server version 3.0 required)

These updates apply to Dragon-based systems only:

• Users can now search for words on the Manage Vocabulary page.
• Users can now filter words on the Manage Vocabulary page based on the level (user/group/site) for which they were added.
• New Modify Word screen: Users can now edit the spelling and pronunciation of words in their personal lexicon.
• Users can now record how a word is pronounced.
• New voice command to record how a word is pronounced: train word.

Personal auto-texts (server version 3.0 required)

These updates apply to Dragon-based and SpeechMagic-based systems:

• Users can now create auto-texts via the personalization and help window, Manage Auto-texts page. This new functionality is available to desktop users and mobile device users with a Dragon Medical One license.
• New voice command: make that an auto-text.

Speech recognition (server version 3.0 required)

These updates apply to Dragon-based systems only:

• New English, Dutch, Flemish and German voice commands to form compound words. English examples: compound, compound that, compound selection.
• New German voice commands to separate compound words: trenne das, Auswahl trennen, trenne Auswahl.
• Improved correction alternatives for Dutch, French and German in the correction menu.
• Dragon Medical now intelligently identifies sentence separators; this means voice commands for selecting sentences and navigating to sentences work more reliably.

These updates apply to Dragon-based and SpeechMagic-based systems:

• New voice commands to navigate text: go to beginning of sentence, go to end of sentence, go to beginning of paragraph, go to end of paragraph.

Android edition

• Support for the Android Marshmallow permission model.
• Dragon Medical SpeechKit performance has been improved in high latency situations.

On-premises systems (server version 3.0 required)

• Dragon Medical systems now remove speech recognition data from the system file share when the data is no longer needed for improving the recognition quality.

Known issues

• With an Android emulator, recording and speech recognition are only supported on Windows platforms. Speech recognition quality might be lower with an emulator than with an actual device. [DNB-6098]

• For SpeechMagic-based systems the SpokenText parameter in the CommandRecognized event is empty in command mode. [DNB-328]

• The speech bar covers part of the soft keyboard if your Activity has its theme set to "android:theme="@android:style/Theme.Dialog" in the AndroidManifest.xml file. For the speech bar to be positioned correctly, you must set windowSoftInputMode to stateHidden in the activity tag. For example:

  <activity
      android:windowSoftInputMode="stateHidden"
      android:theme="@android:style/Theme.Dialog"
      ...>
  </activity>
  

Once this setting is made, the speech-enabled activity will start with the keyboard hidden.

Infrastructure

Infrastructure no longer supported

• Version 3.0 client applications do not work with version 1.5 servers.
• In Dragon Medical Server 3.0 all version 1.4 client-based applications will stop working.

Infrastructure changes planned for the next release

• In Dragon Medical Server 3.1 all version 1.5 and 1.6 client-based applications stop working.
• All applications based on Dragon Medical SpeechKit 3.1 will stop working with 360 | SpeechAnywhere Services 1.6 servers.

System requirements*

*End-of-life: third party environments

Support for third party environments is only valid as long as they are supported by the corresponding vendor and might be subject to other restrictions. Please contact Nuance Technical Support for details. For more information, see the documentation delivered with the third-party product and supporting Nuance documentation.

When support by the vendor has stopped, Nuance will continue support if an issue is specific to the Nuance solution, within the limitations of the vendor’s and Nuance's end-of-life policy. This means that issues that are a combination of the vendor’s component and the Nuance solution cannot be supported.

SpeechAnywhere/Dragon Medical SpeechKit product life cycle

Nuance provides support throughout the life cycle of one version and point release. Please note that Nuance will not perform problem analysis after the end-of-life date stated here. If a Nuance-hosted system is used, Nuance cannot guarantee that an older version will continue working after the date stated here.