Constraint-Based Pitch-Cycle Identification Using a Hybrid Temporal/Spectral Method
Identification of glottal pulses in speech signals allows pitch synchronization during speech coding and recognition. Some applications do not require real time operation, for example automated annotation of speech corpora for use as databases in Text-To-Speech systems. This paper describes a multi-pass, constraints-based method of generating pitch cycle markers for voiced sections of continuous speech. Examples are given that compare the performance of this automated system with other candidate algorithms and pulse identifications made using throat microphone signals.
Click to purchase paper or login as an AES member. If your company or school subscribes to the E-Library then switch to the institutional version. If you are not an AES member and would like to subscribe to the E-Library then Join the AES!
This paper costs $20 for non-members, $5 for AES members and is free for E-Library subscribers.