WTASR: Wavelet Transformer for Automatic Speech Recognition of Indian Languages

Automatic speech recognition systems are developed for translating the signals into corresponding text representation. This translation is used in a variety of applications like voice enabled commands, assistive devices and bots, etc. There significant lack efficient technology Indian languages. In this paper, an wavelet transformer automatic (WTASR) language proposed. The suffer from problem high low frequency over different times due to variation speaker. Thus, wavelets enable network analyze signal multiscale. decomposition fed generating text. comprises encoder decoder system translation. model trained on dataset proposed method compared with other state art methods. results show that WTASR has word error rate can be effective language.