Simulation of an Artificial Hearing Module for an Assistive Robot
| Autor: | J. J. F. Cerqueira, Eduardo F. Simas Filho, Marcio L. L. Oliveira |
|---|---|
| Rok vydání: | 2018 |
| Předmět: |
0209 industrial biotechnology
020901 industrial engineering & automation Artificial neural network Computer science Speech recognition 020208 electrical & electronic engineering Feature extraction 0202 electrical engineering electronic engineering information engineering Assistive robot 02 engineering and technology Mel-frequency cepstrum Multilateration Classifier (UML) |
| Zdroj: | Advances in Intelligent Systems and Computing ISBN: 9783030010539 IntelliSys (1) |
| Popis: | It is proposed in this paper a simulation of an artificial hearing module for a low-cost assistive robot with the purposes of: (1) identifying the speaker; (2) determining the speaker’s location; (3) estimating what emotion is transmitted by his or her speech; and (4) determining an algorithm that links the previous tasks. Mel Frequency Cepstral Coefficients (MFCCs) are used as feature extraction technique, artificial neural networks as classifier and multilateration as technique for locating a sound source. Results from both simulated and public data sets indicate the efficiency of the proposed method. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|