Arabic Speech Recognition Independent of Vocabulary for Isolated Words

This paper presents a vocabulary-independent speech recognition model of isolated Arabic words. This approach uses a restricted dictionary of syllables and phonemes, and it is based on the hidden Markov model (HMM). The main objective of this contribution is to remedy the problem of insufficient vocabularies in automatic speech recognition systems. This new model has made it possible to recognize any given word even if it has never been pronounced. The model gives a very considerable recognition rate comparable to the limited vocabulary speech recognition system of isolated words. © 2022, The Author(s), under exclusive license to Springer Nature Singapore Pte Ltd.