A Comprehensive Review of Automatic Speech Recognition Systems

Speech remains the most effortless and natural way for humans to connect. Also new technological advancements have made humans to not only interact with other humans but also with machines nowadays. In recent years, there has been a tremendous amount of study into speech signal processing. Particularly, the concept of Automated Speech Recognition (ASR) technologies has experienced an upward trend in interest. ASR began off as basic systems that could only recognize a few numbers of sounds and have since grown into complex systems that can understand and speak genuine language. This comprehensive analysis of ASR is offered to assist other scholars with the most important subjects covered in recent years. Additionally, this research will assist in identifying current significant ASR difficulties in practical settings. It also talks about the shortcomings in ASR research now. The accuracy and performance of the various ASR models used for speech recognition are discussed in this paper. The findings presented in this review shed insight on the ASR models that perform the best.