Real-Time Speech Signal Segmentation Methods

We present two algorithms developed for the purpose of real-time speech signal analysis. The first algorithm's function is for the vowel region detection and the second is the rate of speech estimator. Both algorithms have many applications in the field of speech processing, e.g., automatic speech recognition, automatic language identification, automatic emotion recognition, etc. Evaluation of the proposed algorithms assessed their accuracy, reliability, and real-time operating capabilities and our experimental results show that the proposed algorithms perform equally well or better than the commonly known offline approaches.