Learn More
This paper presents a new spectral envelope conversion method using deep neural networks (DNNs). The conventional joint density Gaussian mixture model (JDGMM) based spectral conversion methods perform stably and effectively. However, the speech generated by these methods suffer severe quality degradation due to the following two factors: 1) inadequacy of(More)
The spectral envelope is the most natural representation of speech signal. But in voice conversion, it is difficult to directly model the raw spectral envelope space, which is high dimensional and strongly cross-dimensional correlated, with conventional Gaussian distributions. Bidirectional associative memory (BAM) is a two-layer feedback neural network(More)
The basic assumption in the prediction of peak ground acceleration (PGA), peak ground velocity (PGV), and pseudoresponse spectral values by the random vibration theory (RVT) method is that the ground motion process is a band-limited Gaussian random process (BGRP). However, for the estimation of pseudo-response spectral values, the process is the output of a(More)
AIM To observe the difference of EMG features between junior athletes and general students of knee flexor and extensor in performing various jumps. METHODS 30 junior athletes and 30 middle school students took part in the test. EMG signal from knee flexor and extensor were measured, when subjects performed squatting jump, counter-movement jump and drop(More)
  • 1