Speech signal interpolation device, speech signal interpolation method, and program
View Patent ↗A voice signal interpolation apparatus is provided which can restore original human voices from human voices in a compressed state while maintaining a high sound quality. When a voice signal representative of a voice to be interpolated is acquired by a voice data input unit 1 , a pitch deriving unit 2 filters this voice signal to identify a pitch length from the filtering result. A pitch length fixing unit 3 makes the voice signal have a constant time length of a section corresponding to a unit pitch, and generates pitch waveform data. A sub-band dividing unit 4 converts the pitch waveform data into sub-band data representative of a spectrum. A plurality of sub-band data pieces are averaged by an averaging unit 5 and thereafter a sub-band synthesizing unit 6 converts the sub-band data pieces into a signal representative of a waveform of the voice by a sub-band synthesizing unit 6 . The time length of this signal in each section is restored by a pitch restoring unit 7 and a sound output unit 8 reproduces the sound represented by the signal.
1. A voice signal interpolation apparatus comprising:
pitch waveform signal generating means for acquiring an input voice signal representative of a waveform of voice and making a time length of a section corresponding to a unit pitch of said input voice signal be substantially the same to transform said input voice signal into a pitch waveform signal;
wherein said pitch waveform signal generating means comprises:
a variable filter whose frequency characteristics can be controlled to be variable, said variable filter filtering said input voice signal to derive a fundamental frequency component of the input voice;
filter characteristic determining means for identifying a fundamental frequency of the input voice in accordance with the fundamental frequency component derived by said variable filter and controlling said variable filter so as to have the frequency characteristics cutting off frequency components other than frequency components near the identified fundamental frequency;
wherein said filter characteristic determining means comprises:
cross detecting means for identifying a period of timings at which the fundamental frequency components derived by said variable filter reach a predetermined value and identifying the fundamental frequency in accordance with the identified period;
average pitch detecting means for detecting a time length of a pitch of voice represented by said input voice signal in accordance with said input voice signal before being filtered; and
judging means for judging whether the period identified by said cross detecting means and the time length of the pitch identified by said average pitch detecting means are different from each other by a predetermined amount or more, if it is judged that the period and the time length are not different, controlling said variable filter so as to have the frequency characteristics cutting off frequency components other than frequency components near the fundamental frequency identified by said cross detecting means, and if it is judged that the period and the time length are different, controlling said variable filter so as to have the frequency characteristics cutting off frequency components other than frequency components near a fundamental frequency identified from the time length of the pitch identified by said average pitch detecting means.
2. A voice signal interpolation apparatus according to claim 1 , wherein said voice signal interpolation apparatus comprises:
spectrum deriving means for generating data representative of a spectrum of said input voice signal in accordance with the pitch waveform signal;
averaging means for generating averaged data representative of a spectrum of a distribution of average values of respective spectrum components of said input voice signal, in accordance with a plurality of data pieces generated by said spectrum deriving means; and
voice signal restoring means for generating an output voice signal representative of voice having a spectrum represented by the averaged data generated by said averaging means.
3. A voice signal interpolation apparatus according to claim 1 , wherein said pitch waveform signal generating means comprises: characteristics cutting off frequency components other than frequency components near the identified fundamental frequency;
pitch deriving means for dividing said input voice signal into a voice signal in the section corresponding to the unit pitch, in accordance with a value of the fundamental frequency component derived by said variable filter; and
pitch length fixing means for generating the pitch waveform signal having substantially the same time length in each section by sampling each section of said input voice signal at substantially the same number of samples.
4. A voice signal interpolation apparatus according to claim 1 , wherein said average pitch detecting means comprises:
cepstrum analyzing means for calculating a frequency at which a cepstrum of the input voice signal before filtered by said variable filter takes a maximal value;
self-correlation analyzing means for calculating a frequency at which a periodgram of the input voice signal before filtered by said variable filter takes a maximal value; and
average calculating means for calculating an average value of pitches of voice represented by the input voice signal in accordance with the frequencies calculated by said cepstrum analyzing means and said self-correlation analyzing means and identifying the calculated average value as the time length of the pitch of the voice.