Emotion Text to Speech



Emotion Text to Speech (ETTS) service converts text and emotion values to audio data.


Process of ETTS Engine


ETTS service by LG AI Platform supports the following features:

Feature Description
Comprehensive analysis

Analyzes text inputted by the user, examining pronunciation, syllables, phrases, context, parts of speech, etc.

Advanced speech synthesis

Learns words and writings based on artificial neural network technology, deduce meanings according to a given text, and produce speech data. (based on a Sequence to Sequence model).

Fast response time

Generates audio data as soon as the user inputs text.

Human-like speech

Employs natural, human-like speech using the latest deep learning technology.

Multi-language support

Korean and English

Voice emotion selection

Allows the selection of voice emotion according to user preferences

Voice gender selection

Allows the selection of voice gender according to user preferences

Speech speed control

Offers five-speed levels to deliver speech output.

Pitch tuning

Customizes voice pitch according to user preferences.

Volume control

Offers ten-levels of volume control.



ETTS service passes text entered by the user and emotion data to ETTS Server and stores the audio data passed by ETTS Server in storage.

ETTS Server processes the data passed by ETTS service to generate audio data and passes it to ETTS service.


Architecture of ETTS Engine


Examples of Use

ETTS service is used in a variety of areas that require voice-based services. ETTS service can be also used in audio description for the visually impaired, audiobook, or any application that plays voice audio.


  • Voice of Artificial Intelligence Robot

The robot can communicate with the user with an emotional voice.

  • ARS Guided Voice

ARS guidance service is provided by an emotional voice depending on the customer's mood.

  • Voice on navigation

Navigation informs the destination and direction with emotion and voice set by the user.