pylips.speech.PollyTTS

class pylips.speech.PollyTTS[source]

A text-to-speech backend that uses Amazon Polly.

This class is used to generate audio files from text using Amazon Polly. It can also generate visemes that correspond to the audio files that it generates.

Parameters: None –

__init__()[source]

Methods

`__init__`()
`gen_audio_and_visemes`(text[, voice_id, fname])	Generates an audio file and visemes from a string of text using Amazon Polly.
`get_audio_and_visemes`(fname)	Loads presaved audio and visemes from a file.
`list_voices`()	Lists all the voices that are available in the Amazon Polly TTS backend.

gen_audio_and_visemes(text, voice_id=None, fname=None)[source]

Generates an audio file and visemes from a string of text using Amazon Polly.

Parameters

text (str) – the text that the robot should speak
voice_id (str) – the voice that the robot should speak in
fname (str) – the name of the file that the audio should be saved to

Returns

a tuple containing fname, times, and visemes. fname is the path to the audio file, times is a list of times that correspond to the initiation of the visemes, and visemes is a list of visemes that correspond to the words in the audio

Return type

(tuple)

get_audio_and_visemes(fname)[source]

Loads presaved audio and visemes from a file.

Parameters: fname (str) – the name of the file that the audio and visemes were saved to
Returns: a tuple containing fname, times, and visemes. fname is the path to the audio file, times is a list of times that correspond to the initiation of the visemes, and visemes is a list of visemes that correspond to the words in the audio
Return type: (tuple)
Raises: Exception – if the file does not exist

list_voices()[source]

Lists all the voices that are available in the Amazon Polly TTS backend.

For a more in-depth look at the voices, see the Amazon Polly documentation.

Parameters: None –