Diseño e implementación de un sistema traductor de lengua de señas mediante inteligencia artificial para personas con discapacidad auditiva
This research project called "Design and implementation of a sign language translator system through artificial intelligence for people with hearing disabilities" aims to take advantage of the evolution of technology to reduce the communication barriers between hearing people and people wh...
Na minha lista:
| Autor principal: | |
|---|---|
| Formato: | bachelorThesis |
| Idioma: | spa |
| Publicado em: |
2022
|
| Assuntos: | |
| Acesso em linha: | http://dspace.unach.edu.ec/handle/51000/8817 |
| Tags: |
Adicionar Tag
Sem tags, seja o primeiro a adicionar uma tag!
|
Adicionar Tag | Resumo: | This research project called "Design and implementation of a sign language translator system through artificial intelligence for people with hearing disabilities" aims to take advantage of the evolution of technology to reduce the communication barriers between hearing people and people who suffer from hearing loss. Hearing disability prevents people from interacting with their environment and hinders their development in different educational, work, and social environments. Given the lack of attention to these vulnerable groups, a system is developed that allows sign language translation to speed up the communication process, which is why artificial intelligence is used, one of the technological advances that improve the quality of people's lives. Neural networks were also used that, through prior training, can perform a number of tasks, obtaining results with great precision. To carry out the neural network training, it was necessary to use several tools and algorithms that optimize specific processes and provide quick solutions, one of them being Mediapipe, which facilitates the detection of the hand from the frames captured by the system. This is done thanks to the automatic learning models it has. Each frame is processed to obtain important information about the reference points of the hand, with 21 points being those that can be extracted from the hand and each one made up of coordinates X, Y, and Z that they are the position of each phalanx or knuckle of the hand. This information is later processed and stored to be used in the training of the neural network. Based on the results obtained, it is verified that the implemented system works correctly and is a beneficial tool for people with hearing disabilities. It allows them to improve their communication with a high level of reliability in good lighting conditions; the system detects 94 .46%, while in normal light conditions, it detects 92.08%. Finally, it detects 89.15% of bad light conditions the gestures made. |
|---|