ESCA Workshop on Automatic Speaker Recognition, Identification, and Verification
Some of semiautomatic speaker recognising systems exploit the individual difference of vowels features, such as fundamental and formant frequencies. Linear prediction coefficients can be assessed as well.
These differences are processed by means of multi-varied statistical analysis. Raw data used for analysis are yielded by parameters measures.
The aim of this paper is to evaluate results reliability, when parameters determination is accomplished by human operator supported by commercial software for both voice and statistical analysis.
Bibliographic reference. Giua, Paolo / Pisani, Raffaele (1994): "Intra- and inter-speaker variability", In ASRIV-1994, 123-126.