Paper Contents
Abstract
AbstractSpeech recognition has emerged as one of themost important areas of human-computer interactionthat, with the introduction of NLP methodologies, hasripened into a sophisticated discipline. This paperpresents developments in NLP-based applications thattake a refreshing view on the traditional speech-to-textframework based on phoneme recognition, acousticmodeling, and contextual understanding. Accent, dialect, and noisy environment have been some of thechallenging situations deep learning models liketransformers and recurrent neural networks have takenon. Accent, dialect, noise in surrounding environmentsare covered by present-day means of real-timeprocessing capabilities, support for multiple languages, and semantic exactness in speech recognition. Results ofexperiments reflected functionality of using naturallanguage processing-based speech recognition for manyapplications such as virtual assistants, transcription tools, and accessibility instruments. Going forward, moreenhanced and more inclusive speech recognition systemsare apparent with the application of pre-trainedlanguage models and multimodal data . Key terms: speech recognition, natural languageprocessing, deep learning, acoustic modeling, transformers, phoneme identification, multilingual, semantic precision.I. INTRODUCTION
Copyright
Copyright © 2024 sakshi joshi. This is an open access article distributed under the Creative Commons Attribution License.
