General Objective: Develop an interpretation system for Ecuadorian sign language through the use of artificial intelligence tools based on transfer learning and deep learning in order to improve the communication and interaction of deaf people within the Ecuadorian context.
Specific objectives:
- Collect and annotate an LSEc data set by planning and coordinating video recording sessions with LSEc speakers, annotating collected videos following a protocol and performing quality controls on the data, in this way you can obtain a data set suitable for training and evaluating transfer learning models.
- Develop an AI model for the interpretation of the LSEc using static and moving images of the LSEc, which, together with pre-trained sign language classification models, will allow interaction between speakers of the LSEc and the spoken Spanish language.
- Validate the AI model for the interpretation of LSEc into Spanish with real users, through structured tests that allow measuring both the accuracy of the system and the level of user satisfaction.
Participating Institutions:
UTI, EPN, PUCE, UPS
Project Director Andrès Rubio Proaño
Participants:
- Christian Raul Salamea Palacios
- Xavier Alexander Calderon Hinojosa
- Andres Xavier Rubio Proaño
- Marcelo Javier Sotaminga Cinilín
- Juan Pablo Zaldumbide Proaño
- Jorge Luis Banet Ponce
- Janine Marie Matts
Awarded budget: $39,810.16
Project Status: Awarded
