Sentence examples for modulation spectrum that from inspiring English sources

A good choice for the envelope detector is the incoherent detector, since it is able to create a modulation spectrum that has a large area covered in the modulation frequency domain.

On the contrary, it is now necessary to automatically extract the information encoded in the modulation spectrum that humans would use to understand the message.

We have shown that the output of a modulation spectrum analyser that does not involve any form of conversion to the equivalent of a short-time power spectrogram is able to exploit the spectro-temporal continuity constraints that are typical for speech and which are a prerequisite for noise robust ASR.

In this paper we explore the use of a modulation spectrum front end that is based on time-domain filtering that does not require collapsing the output to the equivalent of one-third octave filters, but which still makes it possible to estimate the posterior probability of the states in a finite-state machine.

The experiments with the MLP classifiers for obtaining state posterior probabilities from the modulation spectrum features confirm that the modulation spectrum features capture most of the information that is relevant for speech decoding.

We need to process that with a low-pass filter that extracts the part of the spectrum around DC and throws away the upper part of the spectrum that gets generated in the second modulation process.

Although cluster purity does not guarantee high recognition performance, from Tables 1 and 2 it can be seen that the modulation spectrum features appear to capture substantial information that can be exploited by two very different classifiers.

The two systems that used modulation spectrum features perform much worse on clean speech than the MLP-based system that used nine adjacent 10-ms PLP +Δ+Δ Δ features [37].

This suggests that the modulation spectrum features fail to capture part of the dynamic information that is represented by the speed and acceleration features derived from PLPs.

This suggests that our modulation spectrum features are suboptimal with respect to describing the medium-term dynamics of the speech signal.

Indeed, from a comparison with the results obtained with the original SC system using five-frame stacks in [24], it appears that the modulation spectrum features outperform stacks of five Mel-spectrum features in all but one condition.