In this work, CSPHMM2s have been proposed, implemented and evaluated to enhance speaker identification performance in the shouted/angry talking environments.
This figure indicates that increasing the value of α has a significant impact on enhancing speaker identification performance in the shouted talking environments.
In four of his earlier studies, Shahin focused on enhancing speaker identification performance in shouted talking environments based on each of Second-Order Hidden Markov Models (HMM2s) [8], Second-Order Circular Hidden Markov Models (CHMM2s) [9], Suprasegmental Hidden Markov Models (SPHMMs) [10], and gender-dependent approach using SPHMMs [11].
This would include having stated objectives or summarizing material during the lecture both of which have been felt to enhance speaker effectiveness [ 1].
It is important to note that the paradigm employed here was special in making the speaker's identity known to the listener before each block of test trials, thus giving ample opportunity for listeners to mentalize and access common ground to enhance speaker-specific linguistic predictions.
But, here, the speaker activity detector presented in Appendix Appendix 2: enhanced fullband speaker activity detection based on multipath-induced fading patterns is used, and instead of processing a time domain signal for peak tracking, a root-mean-square measure over all subbands is applied.
For further details, please refer to Appendix Appendix 2: enhanced fullband speaker activity detection based on multipath-induced fading patterns.
Now, we evaluate the fullband SAD introduced in Section 6.1 and outlined further in Appendix Appendix 2: enhanced fullband speaker activity detection based on multipath-induced fading patterns.
For details concerning the robust fullband SAD detector SAD ̂ m , please refer to Section 6.1 and Appendix Appendix 2: enhanced fullband speaker activity detection based on multipath-induced fading patterns.
