Direct Recovery of Clean Speech Using a Hybrid Noise Suppression Algorithm for Robust Speech Recognition System

<div class="floats-full-page floats-full-page-other"><h2>Table 5</h2><div class="floats-full-page-caption">Recognition results for comparison targets.</div><div class="floats-full-page-content"><table class="table-group" id="tab5"><tr><td><table class="table"><tr><td class="thead-hr" colspan="5"><hr /></td></tr><tr class="thead"><td align="left"> </td><td align="center">Avg 0–20</td><td align="center">Relative Imp.</td><td align="center"><span style="width: 56.862499px;"><svg height="11.6875" id="M87" style="vertical-align:-0.20473pt" version="1.1" viewbox="0 0 56.862499 11.6875" width="56.862499" xmlns="http://www.w3.org/2000/svg" xmlns:xlink="http://www.w3.org/1999/xlink">
<g transform="matrix(.017,-0,0,-.017,.062,11.363)"><path d="M409 504l-29 -5q-16 60 -44.5 96t-86.5 36q-54 0 -82.5 -32t-28.5 -77q0 -51 30.5 -82.5t96.5 -65.5l35.5 -18t33 -19.5t33.5 -23.5l27 -25.5t24.5 -32t13.5 -36.5t6 -45q0 -80 -62 -134.5t-160 -54.5q-47 0 -98 15q-22 7 -50 21q-8 23 -27 155l30 7q7 -27 18 -52&#10;t30 -51.5t49 -42.5t67 -16q56 0 88 32.5t32 87.5q0 51 -31.5 82t-98.5 67q-80 44 -110 73q-55 53 -55 124q0 75 56 126.5t150 51.5q53 0 126 -23z" id="x53"></path></g><g transform="matrix(.017,-0,0,-.017,7.916,11.363)"><path d="M719 650v-28q-43 -2 -62 -15t-22 -44q-6 -47 -6 -169v-403h-31l-426 524h-2v-251q0 -111 6 -169q4 -37 24 -50.5t72 -16.5v-28h-237v28q45 2 64.5 16t23.5 49q6 62 6 171v220q0 54 -3 68.5t-17 32.5q-16 19 -34.5 26.5t-54.5 10.5v28h147l418 -502h3v246q0 117 -7 166&#10;q-4 34 -24.5 47t-73.5 15v28h236z" id="x4E"></path></g><g transform="matrix(.017,-0,0,-.017,20.546,11.363)"><path d="M631 18l-4 -26q-24 0 -35 1q-57 3 -92.5 23t-69.5 68q-21 30 -101 160q-13 22 -30.5 31t-54.5 9h-32v-159q0 -63 14 -77.5t74 -19.5v-28h-262v28q62 5 76 19.5t14 77.5v402q0 62 -13 76t-73 19v28h248q109 0 159 -34q67 -42 67 -131q0 -115 -127 -172q27 -52 89 -144&#10;q49 -71 81 -104q31 -36 72 -47zM212 316h48q73 0 108 29q55 42 55 124q0 76 -42 112t-104 36q-41 0 -54 -10q-11 -7 -11 -44v-247z" id="x52"></path></g><g transform="matrix(.017,-0,0,-.017,34.876,11.363)"><path d="M535 230h-483v50h483v-50z" id="x2212"></path></g><g transform="matrix(.017,-0,0,-.017,48.628,11.363)"><path d="M153 550l-26 -186q79 31 111 31q90 0 141.5 -51t51.5 -119q0 -93 -89 -166q-85 -69 -173 -71q-32 0 -61.5 11.5t-41.5 23.5q-18 17 -17 34q2 16 22 33q14 9 26 -1q61 -50 124 -50q60 0 93 43.5t33 104.5q0 69 -41.5 110t-121.5 41q-53 0 -102 -20l38 305h286l6 -8&#10;l-26 -65h-233z" id="x35"></path></g>
</svg></span></td><td align="center">Relative Imp.</td></tr><tr><td class="thead-hr" colspan="5"><hr /></td></tr><tr><td align="left">MFCC</td><td align="center"><b>71.80</b></td><td align="center">19.4%</td><td align="center"><b>13.39</b></td><td align="center">108.0%</td></tr><tr><td align="left">SS</td><td align="center"><b>72.42</b></td><td align="center">18.4%</td><td align="center"><b>25.26</b></td><td align="center">10.3%</td></tr><tr><td align="left">CMVN</td><td align="center"><b>77.78</b></td><td align="center">10.3%</td><td align="center"><b>13.90</b></td><td align="center">100.4%</td></tr><tr><td align="left">STSA</td><td align="center"><b>80.12</b></td><td align="center">7.0%</td><td align="center"><b>20.31</b></td><td align="center">37.1%</td></tr><tr><td align="left">AFE</td><td align="center"><b>82.23</b></td><td align="center">4.3%</td><td align="center"><b>24.77</b></td><td align="center">12.4%</td></tr><tr><td align="left">MVA</td><td align="center"><b>84.15</b></td><td align="center">1.9%</td><td align="center"><b>26.24</b></td><td align="center">6.1%</td></tr><tr class="table-tr"><td colspan="5"><hr class="tbody-hr" /></td></tr></table></td></tr></table></div></div>

Table 5 | Direct Recovery of Clean Speech Using a Hybrid Noise Suppression Algorithm for Robust Speech Recognition System

title

Table 5: Direct Recovery of Clean Speech Using a Hybrid Noise Suppression Algorithm for Robust Speech Recognition System

description

Table 5

About this journal

Journal profile

Meet the editorial board

Editor spotlight

Find out more

This journal's articles appear in a wide range of abstracting and indexing databases, and are covered by numerous other services that aid discovery and access. Find out more about where and how the content of this journal is available.

International Scholarly Research Notices

Direct Recovery of Clean Speech Using a Hybrid Noise Suppression Algorithm for Robust Speech Recognition System

Table 5