Speaker identification system using peak value envelop lines of vocal waveforms
AssigneeFuji Xerox Co., Ltd.
InventorsMatsumi Suzuki, Hiroyuki Kataoka, Keiichi Ueno
Life of this patent
Estimating the term…
Abstract
Voice verfication of individuals is carried out by utilizing the peak value envelope of voice signals. The voice signal is separated into its positive and negative polarity portions and stepped peak value envelope waveforms are developed from each portion. The latter waveforms are quantized and normalized; the latter being carried out to provide waveforms having a total area therebelow which equals a predetermined area. The quantized, normalized envelopes are then compared with recorded envelopes which were previously calculated in the same manner as the former envelopes. Comparison is carried out by calculating the Hamming distance between the new and stored envelopes.
Claims3
A voice recognition system for developing from an electrical voice signal a characteristic pattern for comparison with a stored characteristic pattern, said system comprising, means for separating said voice signal into its positive and negative polarity portions, peak follower circuit means for following the peaks of said positive and negative polarity portions, sampling means connected to said peak follower circuit means for sampling said peaks at predetermined intervals to develope positive and negative peak envelope waves, and quantizing and normalizing said positive and negative peak envelope waves, said quantized normalized positive and negative peak envelopes constituting said characteristic pattern.
Dependencies are inferred from the claim text.