Auditory Vocoder Demonstration

Paper

• Toshio Irino, Roy D. Patterson, and Hideki Kawahara, "Speech segregation using an auditory vocoder with event-synchronous enhancements," IEEE Trans. Audio, Speech, and Language Process, 14(6), 2212-2221, Nov. 2006.

Examples of segregated sounds

Bable noise (SNR=0dB) (Fig. 8)	Mix 'iradachi' bable	AVoc (0) 'chiNage' bable	AVoc (5) 'doriburu' bable	Comb (0) 'shiNgata' bable	Comb (5) 'heNkaN' bable
Concurrent speaker (SNR=0dB) (Fig. 9)	Mix 'rikujo' 'hitogara'	AVoc (0) 'yuuzuu' 'chirigami'	AVoc (5) 'reNsoo' 'kiNpira'	Comb (0) 'zeNshiN' 'gaNmeN'	Comb (5) 'aojiso' 'zainichi'

Process (F0 error in %). Blue : extracted speech, Green : suppressed speech.

These sounds were extracted from the data set used in the perceptual experiments. Japanese 4-morae words.