# US 11,308,951 B2 · Information processing apparatus, information processing method, and program > Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt - Number: US 11,308,951 B2 - Document ID: US11308951B2 - Status: Granted - Kind: B2 - Application number: 16/471,101 - Filed: 2018-01-04 - Published: 2020-01-16 - Granted: 2022-04-19 - Earliest claimed: 2017-01-18 - Assignees: Sony Corporation - Applicants: Sony Corporation - Inventors: Shinichi Kawano; Yuhei Taki - Web page: https://patentgrep.com/patent/US11308951B2 ## CPC - G10L 15/22: Procedures used during a speech recognition process, e.g. man-machine dialogue - G06F 3/16: Sound input; Sound output (speech processing G10L) - G10L 15/083: {Recognition networks (G10L15/142, G10L15/16 take precedence)} - G10L 15/1815: {Semantic context, e.g. disambiguation of the recognition hypotheses based on word meaning} - G10L 15/26: Speech to text systems (G10L15/08 takes precedence) - G10L 2015/088: {Word spotting} - G10L 2015/223: {Execution procedure of a spoken command} - G10L 2015/228: {of application context} - G10L: SPEECH ANALYSIS TECHNIQUES OR SPEECH SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING TECHNIQUES; SPEECH OR AUDIO CODING OR DECODING - G06F: ELECTRIC DIGITAL DATA PROCESSING (computer systems based on specific computational models G06N) ## Abstract There is provided an information processing apparatus, an information processing method, and a program capable of providing a more convenient speech recognition service. The processing of recognizing, as an edited portion, a desired word configuring a sentence presented to a user as a speech recognition result, acquiring speech information repeatedly uttered for editing a word of the edited portion, and connecting speech information other than a repeated utterance to the speech information is performed, and speech information for speech recognition for editing is generated. Then, speech recognition is performed on the generated speech information for speech recognition for editing. ## Claims ### Claim 1 An information processing apparatus, comprising: a recognition unit configured to recognize a string of words configuring a sentence presented to a user as a speech recognition result, and specify a correspondence relationship between each word of the string of words and phoneme information; an editing correspondence processing unit configured to specify, as an edited portion, a desired word of the string of words as an object to be edited, and delete the phoneme information of the desired word; a generating unit configured to acquire speech information repeatedly uttered for replacing the desired word of the edited portion, and perform processing of connecting a portion of speech information of the speech recognition result other than the word of the edited portion where the phoneme information associated with the desired word is deleted to the speech information of the repeated utterance, to generate speech information for speech recognition for editing; and a speech recognition unit configured to perform speech recognition on the speech information for speech recognition for editing generated by the generating unit, wherein the recognition unit, the editing correspondence processing unit, the generating unit, and the speech recognition unit are each implemented via at least one processor. ### Claim 2 (depends on 1) The information processing apparatus according to claim 1, further comprising a replacement information acquiring unit configured to delete the speech information other than the speech information repeatedly uttered from the speech recognition result obtained by performing speech recognition on the speech information for speech recognition for editing by the speech recognition unit, and acquire the speech information of the repeated utterance as replacement information to replace the edited portion, wherein the replacement information acquiring unit is implemented via at least one processor. ### Claim 3 (depends on 1) The information processing apparatus according to claim 1, wherein the generating unit is further configured to use, as the speech information other than the speech information repeatedly uttered, speech information other than an object to be edited, in which phoneme information corresponding to a word of the edited portion is deleted from speech information representing an entirety of the speech recognition result. ### Claim 4 (depends on 3) The information processing apparatus according to claim 3, wherein the generating unit is further configured to add specific speech information before the edited portion in the speech information other than the object to be edited, and perform processing of connecting the speech information of the repeated utterance following the specific speech information, to generate the speech information for speech recognition for editing. ### Claim 5 (depends on 4) The information processing apparatus according to claim 4, wherein the generating unit is further configured to add the specific speech information in accordance with a part of speech of the word of the edited portion. ### Claim 6 (depends on 4) The information processing apparatus according to claim 4, wherein in accordance with a category to which the word of the edited portion belongs, the generating unit is further configured to add a word, which is easily attached to the category, as the specific speech information. ### Claim 7 (depends on 1) The information processing apparatus according to claim 1, wherein in accordance with a category to which the word of the edited portion belongs, the generating unit is further configured to use synthetic speech representing a sentence generated from a model of the sentence associated with the category, to generate the speech information for speech recognition for editing. ### Claim 8 (depends on 1) The information processing apparatus according to claim 1, wherein the speech recognition unit is further configured to perform speech recognition of an utterance of the user, to generate the sentence presented to the user as the speech recognition result. ### Claim 9 (depends on 1) The information processing apparatus according to claim 1, further comprising: an editing correspondence processing unit configured to delete the word of the edited portion from the sentence presented to the user as the speech recognition result, wherein the generating unit is further configured to perform processing of connecting the speech information repeatedly uttered to a portion of the sentence presented to the user as the speech recognition result where the word of the edited portion is deleted, to generate the speech information for speech recognition for editing, and wherein the editing correspondence processing unit is implemented via at least one processor. ### Claim 10 An information processing method, comprising: recognizing a string of words configuring a sentence presented to a user as a speech recognition result; specifying a correspondence relationship between each word of the string of words and phoneme information; specifying, as an edited portion, a desired word of the string of words as an object to be edited; deleting the phoneme information of the desired word; acquiring speech information repeatedly uttered for replacing the desired word of the edited portion and performing processing of connecting a portion of speech information of the speech recognition result other than the word of the edited portion where the phoneme information associated with the desired word is deleted to the speech information of the repeated utterance, to generate speech information for speech recognition for editing; and performing speech recognition on the generated speech information for speech recognition for editing. ### Claim 11 A non-transitory computer-readable medium having embodied thereon a program, which when executed by a computer causes the computer to execute an information processing method, the method comprising: recognizing a string of words configuring a sentence presented to a user as a speech recognition result; specifying a correspondence relationship between each word of the string of words and phoneme information; specifying, as an edited portion, a desired word of the string of words as an object to be edited; deleting the phoneme information of the desired word; acquiring speech information repeatedly uttered for replacing the desired word of the edited portion and performing processing of connecting a portion of speech information of the speech recognition result other than the word of the edited portion where the phoneme information associated with the desired word is deleted to the speech information of the repeated utterance, to generate speech information for speech recognition for editing; and performing speech recognition on the generated speech information for speech recognition for editing.