In Brazil, Libras is recognized as the mother tongue of the hearing impaired and, despite having a complete linguistic structure, the low adherence to language learning continues to be an impasse for the hearing impaired community in the country. In this context, the use of computational tools that help in the communication process during the literacy of deaf people has shown to be a promising solution. The objective of this work is to develop an application that serves as a pedagogical resource to assist in the communication process between speakers of the Portuguese language and Libras in the literacy process. The methodology consists, firstly, in a systematic mapping, in order to compose the theoretical foundation on the current scenario of Machine Learning focused on the recognition of signs in Libras and define an initial proposal for the software. From this, the construction of the application was carried out, based on the Incremental Model. Finally, an evaluation of the TalkLibras mobile application was carried out, with the acceptance results by the teachers who work with the interested public.
|
