A neural network model for encoding and perception of vowel sounds
β Scribed by Osamu Hoshino; Masayuki Miyamoto; MeiHong Zheng; Kazuharu Kuroiwa
- Book ID
- 114296188
- Publisher
- Elsevier Science
- Year
- 2002
- Tongue
- English
- Weight
- 149 KB
- Volume
- 44-46
- Category
- Article
- ISSN
- 0925-2312
No coin nor oath required. For personal study only.
π SIMILAR VOLUMES
A self-organising neural network has been developed which maps the image velocities of rigid objects, moving in the fronto-parallel plane, topologically over a neural layer. The input is information in the Fourier domain about the spatial components of the image. The computation performed by the net
Neural networks were trained to classify single 20 ms frames of vowels using either perceptually-based spectral representations or LPC spectra as input. Classification performance was compared with performance of several distance measures using nearest-neighbor and mean-distance decision criteria. T