Visualization of Babble–Speech Interactions Using Andrews Curves

Abstract

Visualizing multidimensional data such as the mel frequency cepstral coefficients (MFCCs) proves difficult, especially when the number of dimensions is greater than 3. As a result, it becomes extremely difficult to spot trends in high-dimensional signal interactions. Andrews curves seam to aid in the process of performing graphical analysis of high-dimensional data. This study examines the properties of the babble in the feature domain as well as the effect of the babble noise on the MFCCs of clean speech. Experiments have been conducted using two babble models: the overlapping conversation model and the overlapping speaker model. The purpose of this paper was to provide an insight into the effect of the babble noise on the first thirteen MFCCs of clean speech through the use of Andrews curves. The investigations of this paper give a visual comparison of the signals to expose trends, which the conventional visualization methods do not. The use of Andrews curves not only allows the signal to be observed, but also allows for a statistical comparisons between signals. With a better understanding of the difference between the models, it would be possible to develop systems, which are more robust in babble-corrupted environment.

Fault classification based on computer vision for steel wire ropes

“…After obtaining the feature description vectors of all pictures, the 300 values of each feature are normalized to fall within the range of 0-1. To visualize the threedimension feature descriptors, use Andrews curves [13], which map the data from the highdimensional space onto curves in 2-dimensional space. Figure 6 (a) shows that the Andrews curves of different types differentiate with each other, which means that the proposed feature descriptor can optimally distinguish different types of SWR surface image.…”

Application of static gesture segmentation based on an improved canny operator

“…However, this filtering method also weakens the edges and reduces the edge gradient while optimising the image. Traditional median filtering and mean filtering also have this problem [8–14].…”

“…Vision‐based gesture recognition is the abstraction of various information features of the handle from gestures, which are widely used in various human−computer interaction systems [1–14]. Among them, the most common way to extract gestures is to extract the contour of the hand to describe the characteristics of the hand posture, shape position, and so on.…”