Voice retrieval apparatus, voice retrieval method, and non-transitory recording medium

A voice retrieval method executed by a voice retrieval apparatus that includes a memory storing a sound signal subjected to retrieval, the method comprising:

• a converting step of converting a retrieval string into a phoneme string;
• a time length obtaining step of obtaining, from a database that stores continuous time length data on a phoneme, a continuous time length for each phoneme contained in the phoneme string converted in the converting step;
• an utterance rate information obtaining step of obtaining information on an utterance rate;
• a time length changing step of changing the continuous time length for each of the phonemes obtained in the time length obtaining step in accordance with the information on the utterance rate obtained in the utterance rate information obtaining step;
• a time length deriving step of deriving, based on the continuous time length for each of the phonemes changed in the time length changing step, an utterance time length of voices corresponding to the retrieval string;
• a segment specifying step of specifying, as likelihood obtainment segments, a plurality of segments of the utterance time length derived in the time length deriving step in the sound signal subjected to retrieval;
• a likelihood obtaining step of obtaining a likelihood showing a plausibility that the likelihood obtainment segment specified in the segment specifying step is a segment where voices corresponding to the retrieval string are uttered;
• an identifying step of identifying, based on the likelihood obtained in the likelihood obtainment step, an estimation segment where, in the sound signal subjected to retrieval, utterance of voices corresponding the retrieval string is estimated, the estimation segment being identified for each of the likelihood obtainment segments specified in the segment specifying step;
• and
• a displaying step of displaying on a screen of an output device, the estimation segment identified in the identifying step,
• wherein in the time length changing step, a continuous time length for each vowel phoneme and a continuous time length for each consonant phoneme in the continuous time length for each of the phonemes obtained in the time length obtaining step are changed at respective change rates.

A non-transitory recording medium having recorded therein a program that causes a computer of a voice retrieval apparatus including a memory to record a sound signal subjected to retrieval to execute:

• a time length obtaining process of obtaining, from a database that stores continuous time length data on a phoneme, a continuous time length for each phoneme contained in a phoneme string of a retrieval string;
• an utterance rate information obtaining process of obtaining information on an utterance rate;
• a time length changing process of changing the continuous time length for each of the phonemes obtained in the time length obtaining process in accordance with the information on the utterance rate obtained in the utterance rate information obtaining process;
• a time length deriving process of deriving, based on the continuous time length for each of the phonemes changed in the time length changing process, an utterance time length of voices corresponding to the retrieval string;
• a segment specifying process of specifying, as likelihood obtainment segments, a plurality of segments of the utterance time length derived in the time length deriving process in the sound signal subjected to retrieval;
• a likelihood obtaining process of obtaining a likelihood showing a plausibility that the likelihood obtainment segment specified in the segment specifying process is a segment where voices corresponding to the retrieval string are uttered;
• an identifying process of identifying, based on the likelihood obtained in the likelihood obtainment process, an estimation segment where, within the sound signal subjected to retrieval, utterance of voices corresponding the retrieval string is estimated, the estimation segment being identified for each likelihood obtainment segment specified in the segment specifying process;
• and
• a displaying process of displaying, on a screen of an output device, the estimation segment identified in the identifying process,
• wherein in the time length changing process, a continuous time length for each vowel phoneme and a continuous time length for each consonant phoneme in the continuous time length for each of the phonemes obtained in the time length obtaining process are changed at respective change rates.

A voice retrieval apparatus comprising:

• a processor;
• a memory that records a sound signal subjected to retrieval;
• and
• an output device comprising a screen,
• wherein the processor executes following processes:
• a time length obtaining process of obtaining, from a database that stores continuous time length data on a phoneme, a continuous time length for each phoneme contained in a phoneme string of a retrieval string;
• an utterance rate information obtaining process of obtaining information on an utterance rate;
• a time length changing process of changing the continuous time length for the phonemes obtained in the time length obtaining process in accordance with the information on the utterance rate obtained in the utterance rate information obtaining process;
• a time length deriving process of deriving, based on the continuous time length for each of the phonemes changed in the time length changing process, an utterance time length of voices corresponding to the retrieval string;
• a segment specifying process of specifying, as likelihood obtainment segments, a plurality of segments of the utterance time length derived in the time length deriving process in the sound signal subjected to retrieval;
• a likelihood obtaining process of obtaining a likelihood showing a plausibility that the likelihood obtainment segment specified in the segment specifying process is a segment where voices corresponding to the retrieval string are uttered;
• an identifying process of identifying, based on the likelihood obtained in the likelihood obtainment process, an estimation segment where, within the sound signal subjected to retrieval, utterance of voices corresponding the retrieval string is estimated, the estimation segment being identified for each of the likelihood obtainment segments specified in the segment specifying process;
• and
• a displaying process of displaying on the screen of the output device, the estimation segment identified in the identifying process,
• wherein in the time length changing process, only a continuous time length for each vowel phoneme in the continuous time lengths for respective vowel and consonant phonemes obtained in the time length obtaining process is changed.

A voice retrieval method executed by a voice retrieval apparatus that includes a memory storing a sound signal subjected to retrieval, the method comprising:

• a converting step of converting a retrieval string into a phoneme string;
• a time length obtaining step of obtaining, from a database that stores continuous time length data on a phoneme, a continuous time length for each phoneme contained in the phoneme string converted in the converting step;
• an utterance rate information obtaining step of obtaining information on an utterance rate;
• a time length changing step of changing the continuous time length for the phonemes obtained in the time length obtaining step in accordance with the information on the utterance rate obtained in the utterance rate information obtaining step;
• a time length deriving step of deriving, based on the continuous time length for each of the phonemes changed in the time length changing step, an utterance time length of voices corresponding to the retrieval string;
• a segment specifying step of specifying, as likelihood obtainment segments, a plurality of segments of the utterance time length derived in the time length deriving step in the sound signal subjected to retrieval;
• a likelihood obtaining step of obtaining a likelihood showing a plausibility that the likelihood obtainment segment specified in the segment specifying step is a segment where voices corresponding to the retrieval string are uttered;
• an identifying step of identifying, based on the likelihood obtained in the likelihood obtainment step, an estimation segment where, in the sound signal subjected to retrieval, utterance of voices corresponding the retrieval string is estimated, the estimation segment being identified for each of the likelihood obtainment segments specified in the segment specifying step;
• and
• a displaying step of displaying on a screen of an output device, the estimation segment identified in the identifying step,
• wherein in the time length changing step, only a continuous time length for each vowel phoneme in the continuous time lengths for respective vowel and consonant phonemes obtained in the time length obtaining step is changed.

A non-transitory recording medium having recorded therein a program that causes a computer of a voice retrieval apparatus including a memory to record a sound signal subjected to retrieval to execute:

• a time length obtaining process of obtaining, from a database that stores continuous time length data on a phoneme, a continuous time length for each phoneme contained in a phoneme string of a retrieval string;
• an utterance rate information obtaining process of obtaining information on an utterance rate;
• a time length changing process of changing the continuous time length for the phonemes obtained in the time length obtaining process in accordance with the information on the utterance rate obtained in the utterance rate information obtaining process;
• a time length deriving process of deriving, based on the continuous time length for each of the phonemes changed in the time length changing process, an utterance time length of voices corresponding to the retrieval string;
• a segment specifying process of specifying, as likelihood obtainment segments, a plurality of segments of the utterance time length derived in the time length deriving process in the sound signal subjected to retrieval;
• a likelihood obtaining process of obtaining a likelihood showing a plausibility that the likelihood obtainment segment specified in the segment specifying process is a segment where voices corresponding to the retrieval string are uttered;
• an identifying process of identifying, based on the likelihood obtained in the likelihood obtainment process, an estimation segment where, within the sound signal subjected to retrieval, utterance of voices corresponding the retrieval string is estimated, the estimation segment being identified for each likelihood obtainment segment specified in the segment specifying process;
• and
• a displaying process of displaying, on a screen of an output device, the estimation segment identified in the identifying process,
• wherein in the time length changing process, only a continuous time length for each vowel phoneme in the continuous time lengths for respective vowel and consonant phonemes obtained in the time length obtaining process is changed.