價格：免費
更新日期：2019-07-20
檔案大小：4.0M
目前版本：2.2.0
版本需求：Android 5.0 以上版本
官方網站：http://www.hear2read.org
Email：feedback@hear2read.org
Hear2Read Ⓡ Text-To-Speech (TTS) service engine is for use with all Hear2Read voices.
Please add one or more Hear2Read voices using ADD VOICES button on the App’s Home page after installing this Engine App.
MALAYALAM TTS voice in now available to Beta testers. Please send an e-mail to feedback@Hear2Read.org, if you want to be Beta tester and provide feedback to improve it.
Please read "Using Hear2Read on Android Devices" at https://hear2read.org/solution-android.php for a quick tutorial.
The Engine App runs as a background service to convert text to speech. It supports Android TTS service APIs such that any TTS-enabled App can use it to "Read" by listening to software generated synthetic speech. It does not save the text sent to it or the audio output. All audio output is sent to the App that sent it the text for conversion. Some of these Apps provide an option to save the synthetic speech output.
The App includes a demo screen to let users read by listening to example text. The demo screen is not meant to be a screen reading App.
The App is useful for users with Visual Impairment (VI) with Talkback option under Accessibility in Settings, as well as users that do not have Visual Impairments (aka sighted users).
For Talkback to work effectively, VI users need to set Hear2Read as the preferred TTS Engine under Android “Text-to-speech output” settings and then select the Indic voice they wish to use. In many Android devices, the default voice is "system language" ( most often English). This must be changed to the Indic voice.
Hear2Read engine supports AutoTTS to switch back ad forth between English (ASCII text) and Indic TTS voices. If there is more than one English TTS voice on the device, users can select the voice to speak English text. Numbers are always spoken in the Indic language whether written in English or the Indic script.
When you set Hear2Read as preferred TTS Engine, Android displays the following scary warning message:
"This speech synthesis engine may be able to collect all the text that will be spoken, including personal data like passwords and credit card numbers. ...."
This message is displayed for all third party TTS voices, i.e. TTS voices are not part of the OS provided by the device vendor. It simply means that a TTS engine can save the text sent to it for conversion. Hear2Read does not save or transmit any text read by it.
Some screen reading Apps (such as "@Voice Aloud Reader") determine the content language and then select the appropriate TTS engine and voice. Such Apps do not require users to change the preferred engine to Indic Hear2Read. Such Apps require Talkback setting Audio ducking (Focus Speech Audio) to be turned ON. This allows the screen reading App to pause speaking, when Talkback speaks using the preferred engine, as is the case when the VI users touches any navigation button or home button to open another App.
Some Apps (such as "Voice Dream Reader" for reading eBooks) allow the user to select and associate a TTS Engine and language for each book. This selection is used instead of the preferred engine and voice used by Talkback.
Screen reading Apps are designed to read ASCII (for English) and Unicode (for Indic scripts) text. Most screen reading Apps do not have Optical Character Recognition (OCR ) capability and cannot read scanned .pdf files that contain graphical image. These Apps can read digitally produced .pdf files if they contain Indic text in Unicode. Many older digitally produced .pdf files contain non Unicode text that are not known to these Apps or Hear2Read Engine.
Hear2Read Apps use Festvox, Festival and Flite Open Source software developed by Carnegie Mellon University (CMU) and collaborators including Hear2Read developers and University of Edinburg (UK).
Requires devices with 1 GB (or more) RAM and Quad core CPU.
Please send comments and questions to Feedback@Hear2Read.org.
Note: Hear2Read is a registered trademark.