| Speech Recognition & Synthesis | |
|---|---|
 | |
| Developer | |
| Initial release | 10 October 2013; 12 years ago (2013-10-10) |
| Stable release | |
| Operating system | Android 8+ Discontinued
|
| Type | Screen reader |
Speech Recognition & Synthesis, formerly known asSpeech Services,[3] is ascreen reader application developed byGoogle for itsAndroid operating system. It powers applications to read aloud (speak) the text on the screen, with support for many languages. Text-to-Speech may be used by apps such asGoogle Play Books for reading books aloud,Google Translate for reading aloud translations for the pronunciation of words,Google TalkBack, and other spoken feedback accessibility-based applications, as well as by third-party apps. Users must install voice data for each language.
 | This articleneeds additional citations forverification. Please helpimprove this article byadding citations to reliable sources. Unsourced material may be challenged and removed. Find sources: "Speech Recognition & Synthesis" – news ·newspapers ·books ·scholar ·JSTOR(November 2023) (Learn how and when to remove this message) |
Some app developers have started adapting and tweaking their Android Auto apps to include Text-to-Speech, such asHyundai in 2015.[4] Apps such as textPlus andWhatsApp use Text-to-Speech to read notifications aloud and provide voice-reply functionality.
Google Cloud Text-to-Speech is powered byWaveNet,[5] software created by Google's UK-based AI subsidiaryDeepMind, which was bought by Google in 2014.[6] It tries to distinguish from its competitors,Amazon andMicrosoft.[7]
Most voice synthesizers (including Apple'sSiri) useconcatenative synthesis,[5] in which a program stores individualphonemes and then pieces them together to form words and sentences.WaveNet synthesizes speech with human-like emphasis and inflection on syllables, phonemes, and words.Unlike most other text-to-speech systems, a WaveNet model createsraw audio waveforms from scratch. The model uses a neural network that has been trained using a large volume of speech samples. During training, the network extracts the underlying structure of the speech, such as which tones follow each other and what a realistic speech waveform looks like. When given a text input, the trained WaveNet model can generate the corresponding speech waveforms from scratch, one sample at a time, with up to 24,000 samples per second and smooth transitions between the individual sounds.[5]
The service was renamed Speech Recognition & Synthesis in 2023.[citation needed]
| Software development |
| ||||||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Releases | |||||||||||||||
| Derivatives | |||||||||||||||
| Devices |
| ||||||||||||||
| Custom distributions |
| ||||||||||||||
| Booting and recovery | |||||||||||||||
| APIs | |||||||||||||||
| AlternativeUIs | |||||||||||||||
| Rooting | |||||||||||||||
| Lists | |||||||||||||||
| Related topics | |||||||||||||||
