Uma abordagem de aprendizagem profunda para identificação de pessoas através do som dos passos
With the advent of pervasive computing, the technology has become part of human daily life. Thus, in the construction of ”intelligent”environments, systems are used that are linked to the routine of the individuals who live there. So one of the main needs is the recognition of the individual who liv...
| Autor principal: | Costa, Leonardo Rezende |
|---|---|
| Grau: | Monografia |
| Idioma: | pt_BR |
| Publicado em: |
Universidade Federal do Tocantins
2021
|
| Assuntos: | |
| Acesso em linha: |
http://hdl.handle.net/11612/3196 |
| Resumo: |
|---|
|
With the advent of pervasive computing, the technology has become part of human daily life. Thus, in the construction of ”intelligent”environments, systems are used that are linked to the routine of the individuals who live there. So one of the main needs is the recognition of the individual who lives in the space. The objective of this work is to iden tify individuals in an intelligent environment through information obtained by the sound of their steps. To this end, different configurations of neural networks of deep learning were developed and tested in order to classify the people who participated in an experi ment realized in Carvalho and Rosa (2010). The proposed neural network architecture is composed of a chain of neural networks and the results achieved to 98.57 % accuracy. |