Hear2Read Ⓡ NG Text-To-Speech (TTS) service uses Deep Neural Network (DNN) technology to provide natural (human like) synthetic speech.
This is not an online text to speech App. No internet connection is required to convert text to speech after the Voice (DNN model) for the particular language has been added to the App.
The App is useful for users with Visual Impairment (VI) for use with Talkback, as well as users without Visual Impairments who wish to read by listening.
Speech synthesis using DNN technology requires a text string of several words rather than a character or word at a time. This is necessary to generate natural (not robotic) speech. This results in some delay at the beginning. The delay depends on the CPU speed and the length of the initial text string being synthesized. DNN technology requires faster CPUs than TTS apps that synthesize character or word at a time, such as eSpeak and the original Hear2Read App released in 2016.
Currently it supports the following Indic languages:
* Assamese
* Gujarati
* Hindi
* Kannada
* Malayalam
* Marathi
* Nepali
* Odia
* Punjabi
* Sinhala
* Tamil
* Telugu
After installing the App and the voice(s), set Hear2Read as “Preferred engine” under Text-to-speech output setting.
Hear2Read App includes a demo screen for each supported language. In addition to the prepopulated sample text in the demo screen, users can paste any text (up to 1000 bytes) of their own choosing to read by listening. The demo screen is not a screen reading App.
Hear2Read TTS Engine runs as a background service (same as Google TTS service) to convert text to speech. Any App (such as Talkback or @Voice Aloud) can use it to convert Indic text to speech thereby "Reading by Listening”. It does not save the text sent to it or the speech audio output. All audio output is sent to the App that sent the text for conversion.
Requires devices Android 12 or higher.
Please send comments and questions to Feedback@Hear2Read.org.