Improvement in Automatic Speech Recognition of South Asian Accent Using Transfer Learning of DeepSpeech2

<table class="table-group" id="tab9"><tr><td><table class="table"><tr><td class="thead-hr" colspan="3"><hr/></td></tr><tr class="thead"><td class="align_left">Model</td><td class="align_center">WER (English accent) (%)</td><td class="align_center">WER (non-native English accent) (%)</td></tr><tr><td class="thead-hr" colspan="3"><hr/></td></tr><tr><td class="align_left">DeepSpeech 2</td><td class="align_center">6</td><td class="align_center">43</td></tr><tr><td class="align_left">Proposed</td><td class="align_center">7</td><td class="align_center">18.08</td></tr><tr class="table-tr"><td colspan="3"><hr class="tbody-hr"/></td></tr></table></td></tr></table>

<div>Comparison of DeepSpeech2 and proposed model’s WER on English and non-native English accent.</div>

Mathematical Problems in Engineering

tab9

Table 9

Table 9: Improvement in Automatic Speech Recognition of South Asian Accent Using Transfer Learning of DeepSpeech2 