# US 10,872,597 B2 · Speech synthesis dictionary delivery device, speech synthesis system, and program storage medium > Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt - Number: US 10,872,597 B2 - Document ID: US10872597B2 - Status: Granted - Kind: B2 - Application number: 16/058,229 - Filed: 2018-08-08 - Published: 2019-02-28 - Granted: 2020-12-22 - Earliest claimed: 2017-08-29 - Assignees: Kabushiki Kaisha Toshiba; Toshiba Digital Solutions Cornoration - Applicants: Kabushiki Kaisha Toshiba; Toshiba Digital Solutions Corporation - Inventors: Kouichirou Mori; Gou Hirabayashi; Masahiro Morita; Yamato Ohtani - Web page: https://patentgrep.com/patent/US10872597B2 ## CPC - G10L 13/047: Architecture of speech synthesisers - G10L 13/08: Text analysis or generation of parameters for speech synthesis out of text, e.g. grapheme to phoneme translation, prosody generation or stress or intonation determination - G10L 13/033: Voice editing, e.g. manipulating the voice of the synthesiser - G10L: SPEECH ANALYSIS TECHNIQUES OR SPEECH SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING TECHNIQUES; SPEECH OR AUDIO CODING OR DECODING ## Abstract A speech synthesis dictionary delivery device that delivers a dictionary for performing speech synthesis to terminals, comprises a storage device for speech synthesis dictionary database that stores a first dictionary which includes an acoustic model of a speaker and is associated with identification information of the speaker, that stores a second dictionary which includes an acoustic model generated using voice data of a plurality of speakers, and that stores parameter sets of the speakers to be used with the second dictionary and which are associated with identification information of the speakers, a processor that determines one of the first dictionary and the second dictionary, which should be used in the terminal for a specified speaker, and an input output interface (I/F) that receives the identification information of a speaker transmitted from the terminal and then delivers at least one of a first dictionary, the second dictionary, and a parameter set of the second dictionary, on the basis of the received identification information of the speaker and a result of the determination by the processor. ## Claims ### Claim 1 A speech synthesis dictionary delivery device that delivers a dictionary for performing speech synthesis to a terminal via a network, comprising: a storage device for a speech synthesis dictionary database configured to: store first dictionaries, each of which includes an acoustic model of a speaker and is associated with identification information of the speaker; store a second dictionary that includes a versatile acoustic model generated using voice data of a plurality of speakers; and store parameter sets of the speakers to be used with the second dictionary and that are associated with identification information of the speakers; a processor configured to determine one of a first dictionary and the second dictionary, which should be used in the terminal for a specified speaker, based on a communication state of the network; and an input output interface (I/F) configured to: receive identification information of the specified speaker transmitted from the terminal via the network; and deliver the first dictionary, or at least one of the second dictionary and a parameter set of the second dictionary to the terminal via the network, based on the received identification information of the specified speaker and a result of the determination by the processor. ### Claim 2 (depends on 1) The speech synthesis dictionary delivery device according to claim 1, wherein, after the second dictionary has been transmitted to the terminal, the input output interface is configured to deliver the first dictionary or the parameter set of the second dictionary based on the received identification information of the specified speaker and the result of the determination. ### Claim 3 (depends on 1) The speech synthesis dictionary delivery device according to claim 1, wherein the processor is further configured to: measure the communication state of the network; and determine one of the first dictionary and the second dictionary to be used based on a result of the measurement. ### Claim 4 (depends on 1) The speech synthesis dictionary delivery device according to claim 1, wherein the processor is further configured to: estimate a degree of importance of the specified speaker, and determine one of the first dictionary and the second dictionary to be used based on a result of the estimation. ### Claim 5 (depends on 1) The speech synthesis dictionary delivery device according to claim 1, wherein, when a hardware specification of the terminal is insufficient, the parameter set of the second dictionary is given a priority. ### Claim 6 (depends on 1) The speech synthesis dictionary delivery device according to claim 1, wherein the processor is further configured to: compare acoustic features generated based on the second dictionary with acoustic features extracted from real voice samples of the specified speaker; estimate a degree of reproducibility of a synthesized speech by the second dictionary; and determine one of the first dictionary and the second dictionary to be used based on a result of estimation of the degree of reproducibility. ### Claim 7 A speech synthesis system that delivers a synthetic speech to a terminal via a network, comprising: an input output interface (I/F) configured to receive identification information of a specified speaker transmitted from the terminal via the network; a storage device for a speech synthesis dictionary database configured to: store a first dictionaries, each of which includes an acoustic model of a speaker and is associated with identification information of the speaker; store a second dictionary that includes a versatile acoustic model generated using voice data of a plurality of speakers; and store parameter sets of the speakers to be used with the second dictionary and is associated with identification information of the speakers; a hardware processor configured to: select a first dictionary or a parameter set to be loaded onto the storage device based on a server load of the speech synthesis system; and synthesize a speech using the first dictionary or the parameter set with the second dictionary that is selected by the hardware processor, wherein the input output interface is further configured to deliver the speech synthesized by the hardware processor to the terminal via the network. ### Claim 8 (depends on 7) The speech synthesis system according to claim 7, wherein the hardware processor is further configured to measure the server load of the speech synthesis system, wherein, when the measured server load is not larger than a threshold value, the first dictionary having the lowest usage frequency in loaded ones is unloaded from the storage device, and the first dictionary of the specified speaker requested from the terminal is loaded to the storage device. ### Claim 9 (depends on 7) The speech synthesis system according to claim 7 wherein the hardware processor is further configured to measure the server load of the speech synthesis system, wherein, when the measured server load is larger than a threshold value, the parameter set of the specified speaker requested from the terminal is loaded to the storage device. ### Claim 10 A non-transitory computer-readable storage medium storing instructions that, when executed by one or more processors of a device having a speech synthesis dictionary delivery program stored therein, cause the device to: store first dictionaries each of which includes an acoustic model of a speaker and is associated with identification information of the speaker; store a second dictionary including a versatile acoustic model generated using voice data of a plurality of speakers; store parameter sets of the speakers to be used with the second dictionary in association with identification information of the speakers; determine which of a first dictionary and the second dictionary should be used for a specified speaker based on a communication state of a network connected to a terminal; receive the identification information of the specified speaker transmitted from the terminal via the network; and deliver the first dictionary, or at least one of the second dictionary and a parameter set to the terminal via the network based on the received identification information of the specified speaker and a determination result by the determining. ### Claim 11 A speech synthesis device that provides a synthetic speech to a terminal via the network, comprising: a storage unit for a speech synthesis dictionary database configured to: store first dictionaries each of which includes an acoustic model of a speaker and is associated with identification information of the speaker; store a second dictionary having a versatile acoustic model that is generated using voice data of a plurality of speakers; and store parameter sets of the speakers to be used with the second dictionary in association with identification information of the speakers; a condition determination unit configured to determine which of a first dictionary and the second dictionary should be used for a specified speaker based on a communication state of the network; and a transceiving unit configured to: receive identification information of the specified speaker transmitted from the terminal via the network; and deliver the first dictionary or at least one of the second dictionary and a parameter set of the second dictionary to the terminal via the network based on the received identification information of the specified speaker and a result of the determination by the condition determination unit. ### Claim 12 (depends on 11) The speech synthesis device according to claim 11, wherein, after the second dictionary is transmitted to the terminal, the transceiving unit is further configured to deliver the first dictionary or the parameter set of the second dictionary based on the received identification information of the specified speaker and the result of the determination by the condition determination unit. ### Claim 13 (depends on 11) The speech synthesis device according to claim 11, further comprising: a communication state measuring unit configured to: measure the communication state of the network; and determine which of the first dictionary and the second dictionary should be used based on a result of the measurement. ### Claim 14 (depends on 11) The speech synthesis device according to claim 11, further comprising: a speaker degree-of importance estimation unit configured to: estimate a degree of importance of the specified speaker; and determine which of the first dictionary and the second dictionary should be used based on a result of the estimation. ### Claim 15 (depends on 11) The speech synthesis device according to claim 11, further comprising: a speaker degree-of-reproducibility estimation unit configured to: compare acoustic features generated based on the second dictionary with acoustic features extracted from a real voice of the specified speaker; and estimate a degree of reproducibility of the synthetic speech, wherein the condition determination unit is further configured to determine one of the first dictionary and the second dictionary to be used based on a result of estimation of the degree-of-reproducibility.