sign_language_translator.models.text_to_sign.concatenative_synthesis module

This module defines the ConcatenativeSynthesis class, which represents a rule based model for translating text to sign language.

class sign_language_translator.models.text_to_sign.concatenative_synthesis.ConcatenativeSynthesis(text_language: str | TextLanguage | Enum, sign_language: str | SignLanguage | Enum, sign_format: str | Type[Sign], sign_embedding_model: str | Enum | None = None)

Bases: TextToSignModel

A class representing a Rule-Based model for translating text to sign language by concatenating sign language videos.

property sign_embedding_model: str | None

The name of the model which was used for extracting features from the signs. This name is used in the filenames of the preprocessed signs dataset.

property sign_format: Type[Sign]

The format of the sign language (e.g. slt.Vision.sign.sign.Sign or subclass).

Class that wraps the sign language features e.g. raw videos or landmarks. This class can load the signs from available datasets and concatenate its objects. e.g. slt.Video or slt.Landmarks class.

property sign_language: SignLanguage

An object of slt.languages.sign.SignLanguage class or its child that defines the mapping rules & grammar of a sign language.

property text_language: TextLanguage

An object of slt.languages.text.TextLanguage class or its child that defines preprocessing, tokenization & other NLP functions.

translate(text: str, *args, **kwargs) → Sign

Translate text to sign language.

Parameters:

text – The input text to be translated.

Returns:

The translated sign language sentence.