Busca avançada
Ano de início
Entree


Neural machine translation from text to sign language

Texto completo
Autor(es):
De Martino, Jose Mario ; Silva, Ivani Rodrigues ; Marques, Janice Goncalves Temoteo ; Martins, Antonielle Cantarelli ; Poeta, Enzo Telles ; Christinele, Dener Stassun ; Campos, Joao Pedro Araujo Ferreira
Número total de Autores: 7
Tipo de documento: Artigo Científico
Fonte: UNIVERSAL ACCESS IN THE INFORMATION SOCIETY; v. N/A, p. 14-pg., 2023-07-18.
Resumo

The paper describes an ongoing project aimed at developing a neural machine translation approach to translating text into sign language, with the translation result presented by a realistic three-dimensional avatar. The approach can be applied to translate texts, books, and Internet pages, improving access to information for deaf individuals and contributing to the social, educational, and labor inclusion of these citizens. The paper elaborates on four fundamental issues related to the approach: (i) the establishment of a written representation of the sign language; (ii) the construction of a parallel corpus involving text from an oral language and the written representation of sign language; (iii) the establishment of a neural machine translation model to perform the translation; (iv) the visual presentation of the translation by a signing avatar. Although focused on translating Brazilian Portuguese into Brazilian Sign Language, the concepts discussed in the paper are general enough to be applied to other written-signed language pairs. (AU)

Processo FAPESP: 21/02365-2 - Critérios de lematização para a construção de produtos lexicográficos de Língua de Sinais Brasileira: Libras
Beneficiário:Antonielle Cantarelli Martins
Modalidade de apoio: Bolsas no Brasil - Pós-Doutorado