WO2017080239A1 - 录音标记方法及录音装置 - Google Patents

录音标记方法及录音装置 Download PDF

Info

Publication number
WO2017080239A1
WO2017080239A1 PCT/CN2016/089033 CN2016089033W WO2017080239A1 WO 2017080239 A1 WO2017080239 A1 WO 2017080239A1 CN 2016089033 W CN2016089033 W CN 2016089033W WO 2017080239 A1 WO2017080239 A1 WO 2017080239A1
Authority
WO
WIPO (PCT)
Prior art keywords
voiceprint
recording
target
template
sample
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Ceased
Application number
PCT/CN2016/089033
Other languages
English (en)
French (fr)
Inventor
蔡竹沁
齐峰岩
牛磊
关彬
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Le Holdings Beijing Co Ltd
Lemobile Information Technology (Beijing) Co Ltd
Original Assignee
Le Holdings Beijing Co Ltd
Lemobile Information Technology (Beijing) Co Ltd
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Le Holdings Beijing Co Ltd, Lemobile Information Technology (Beijing) Co Ltd filed Critical Le Holdings Beijing Co Ltd
Publication of WO2017080239A1 publication Critical patent/WO2017080239A1/zh
Anticipated expiration legal-status Critical
Ceased legal-status Critical Current

Links

Images

Classifications

    • G—PHYSICS
    • G10—MUSICAL INSTRUMENTS; ACOUSTICS
    • G10L—SPEECH ANALYSIS TECHNIQUES OR SPEECH SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING TECHNIQUES; SPEECH OR AUDIO CODING OR DECODING
    • G10L15/00—Speech recognition
    • G10L15/02—Feature extraction for speech recognition; Selection of recognition unit
    • G—PHYSICS
    • G10—MUSICAL INSTRUMENTS; ACOUSTICS
    • G10L—SPEECH ANALYSIS TECHNIQUES OR SPEECH SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING TECHNIQUES; SPEECH OR AUDIO CODING OR DECODING
    • G10L15/00—Speech recognition
    • G10L15/06—Creation of reference templates; Training of speech recognition systems, e.g. adaptation to the characteristics of the speaker's voice
    • G—PHYSICS
    • G10—MUSICAL INSTRUMENTS; ACOUSTICS
    • G10L—SPEECH ANALYSIS TECHNIQUES OR SPEECH SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING TECHNIQUES; SPEECH OR AUDIO CODING OR DECODING
    • G10L15/00—Speech recognition
    • G10L15/06—Creation of reference templates; Training of speech recognition systems, e.g. adaptation to the characteristics of the speaker's voice
    • G10L15/063—Training
    • G—PHYSICS
    • G10—MUSICAL INSTRUMENTS; ACOUSTICS
    • G10L—SPEECH ANALYSIS TECHNIQUES OR SPEECH SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING TECHNIQUES; SPEECH OR AUDIO CODING OR DECODING
    • G10L15/00—Speech recognition
    • G10L15/08—Speech classification or search
    • G—PHYSICS
    • G10—MUSICAL INSTRUMENTS; ACOUSTICS
    • G10L—SPEECH ANALYSIS TECHNIQUES OR SPEECH SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING TECHNIQUES; SPEECH OR AUDIO CODING OR DECODING
    • G10L17/00—Speaker identification or verification techniques
    • G—PHYSICS
    • G11—INFORMATION STORAGE
    • G11B—INFORMATION STORAGE BASED ON RELATIVE MOVEMENT BETWEEN RECORD CARRIER AND TRANSDUCER
    • G11B27/00—Editing; Indexing; Addressing; Timing or synchronising; Monitoring; Measuring tape travel
    • G11B27/10—Indexing; Addressing; Timing or synchronising; Measuring tape travel
    • G—PHYSICS
    • G10—MUSICAL INSTRUMENTS; ACOUSTICS
    • G10L—SPEECH ANALYSIS TECHNIQUES OR SPEECH SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING TECHNIQUES; SPEECH OR AUDIO CODING OR DECODING
    • G10L15/00—Speech recognition
    • G10L15/06—Creation of reference templates; Training of speech recognition systems, e.g. adaptation to the characteristics of the speaker's voice
    • G10L15/063—Training
    • G10L2015/0631—Creating reference templates; Clustering

Definitions

  • the embodiments of the present application relate to the field of electronic technologies, and in particular, to a recording and marking method and a recording device.
  • smart phones are gradually integrated into people's daily lives, which not only become daily communication devices, but also become recording devices that are easy to carry everyday.
  • the user can record and save the voice information through the recording application (Application, APP for short) of the smart phone, so that the user can quickly save a piece of voice information that is difficult to directly memorize, and can also use the recording multiple times.
  • Application Application, APP for short
  • the embodiment of the present application provides a recording and marking method and a recording device, which are used to solve the problem of poor convenience and time consuming when manually marking.
  • an embodiment of the present application provides a recording and marking method, including:
  • the target mark information corresponding to the target voiceprint template is obtained from the voiceprint database
  • the current recording is marked using the target tag information.
  • an embodiment of the present application provides a recording apparatus, including:
  • An extraction module configured to collect a current recording and extract a voiceprint feature parameter from the current recording
  • a training module configured to perform voiceprint clustering training on the voiceprint parameter to obtain a target voiceprint template of the voiceprint parameter
  • a determining module configured to determine whether the target voiceprint template is a voiceprint template in a voiceprint database
  • An acquiring module configured to acquire, from the voiceprint database, target tag information corresponding to the target voiceprint template when the determination result of the determining module is YES;
  • a marking module for marking the current recording with the target marking information.
  • embodiments of the present application provide a recording apparatus including a memory, one or more processors, and one or more programs, wherein the one or more programs are executed by the one or more processors Performing the following operations: collecting a current recording and extracting a voiceprint feature parameter from the current recording; performing voiceprint clustering training on the voiceprint parameter to obtain a target voiceprint template of the voiceprint parameter; determining the target Whether the voiceprint template is a voiceprint template in the voiceprint database; if the determination result is yes, obtaining target marker information corresponding to the target voiceprint template from the voiceprint database; using the target marker information The current recording is marked.
  • embodiments of the present application provide a computer readable storage medium having stored thereon computer executable instructions that, in response to execution, cause a recording device to perform an operation, the operation
  • the method includes: collecting a current recording and extracting a voiceprint feature parameter from the current recording; performing voiceprinting on the voiceprint parameter Class training to obtain a target voiceprint template of the voiceprint parameter; determining whether the target voiceprint template is a voiceprint template in a voiceprint database; if the determination result is yes, obtaining the target from the voiceprint database The target mark information corresponding to the voiceprint template; the current record is marked using the target mark information.
  • the recording and marking method and the recording device of the embodiment of the present application obtain the target sound of the voiceprint parameter by collecting the current recording and extracting the voiceprint feature parameters from the current recording, and performing voiceprint clustering training on the voiceprint parameter.
  • a pattern template determining whether the target voiceprint template is a voiceprint template in the voiceprint database, and if the determination result is yes, acquiring target mark information corresponding to the target voiceprint template from the voiceprint database, using the The target mark information marks the current recording.
  • the voiceprint template corresponding to the current recording is identified by the voiceprint, and the marker information corresponding to the current recording is obtained by using the established voiceprint database, thereby marking the current recording, thereby realizing the function of automatically marking the recording, and saving the user. The time when the tag was added.
  • FIG. 1 is a schematic flow chart of a recording mark method according to Embodiment 1 of the present application.
  • FIG. 2 is a schematic diagram of an application example of a recording mark method according to Embodiment 1 of the present application;
  • FIG. 3 is a schematic flowchart of a recording and marking method according to Embodiment 2 of the present application.
  • FIG. 4 is a schematic diagram of an application example of a recording mark method according to Embodiment 2 of the present application.
  • FIG. 5 is a second schematic diagram of an application example of a recording mark method according to Embodiment 1 of the present application.
  • FIG. 6 is a third schematic diagram of an application example of a recording mark method according to Embodiment 1 of the present application.
  • FIG. 7 is a schematic flowchart of a method for establishing a voiceprint database in Embodiment 2 of the present application.
  • FIG. 8 is a schematic structural diagram of a recording apparatus according to Embodiment 3 of the present application.
  • FIG. 9 is a schematic structural diagram of a recording apparatus according to Embodiment 4 of the present application.
  • FIG. 10 is a schematic structural diagram of still another embodiment of a recording apparatus provided by the present application.
  • FIG. 11 is a schematic structural diagram of an embodiment of a computer program product for recording marks provided by the present application.
  • FIG. 1 it is a schematic flowchart of a recording and marking method according to Embodiment 1 of the present application, and the recording marking method includes:
  • Step 101 Acquire a current recording and extract a voiceprint feature parameter from the current recording.
  • the user can open the recording function of the recording APP downloaded in the smart phone through the user interface of the smart phone, and the recording APP starts to collect the current recording.
  • the recording APP can preprocess the sound, for example, the collected data is performed. Framing, windowing, filtering, etc.
  • acoustic characteristic parameters include: energy of the sound, formant, and Mel-frequency cepstrum coefficients (MFCC). And Linear Prediction Coefficients (LPC).
  • MFCC Mel-frequency cepstrum coefficients
  • LPC Linear Prediction Coefficients
  • Step 102 Perform voiceprint clustering on the voiceprint parameter to obtain a target voiceprint template of the voiceprint parameter.
  • a voiceprint clustering trainer is arranged, and after obtaining the voiceprint feature parameters, the voiceprint clustering training is performed on the voiceprint feature parameters by the trainer, and the current The target voiceprint template corresponding to the recording.
  • Step 103 Determine whether the target voiceprint template is a voiceprint template in a voiceprint database.
  • the voice sound clustering training is performed on the sample sound by the trainer, and the sample voiceprint template corresponding to the sample sound is obtained, and a voiceprint database is preset in the recording APP by using the sample voiceprint template.
  • a plurality of sample voiceprint templates are stored in the general voiceprint database, so that the user can perform recording marks during the recording process.
  • the recording APP can search in the voiceprint database to determine whether the target voiceprint template exists in the voiceprint database.
  • Step 104 If the target voiceprint template is a voiceprint template in the voiceprint database, obtain target marker information corresponding to the target voiceprint template from the voiceprint database.
  • the sample voiceprint template In the voiceprint database, not only the sample voiceprint template is stored, but also the marker information corresponding to the sample voiceprint template is stored. Generally, each sample voiceprint template corresponds to the corresponding marker information.
  • the target marker information corresponding to the target voiceprint template may be acquired.
  • Step 105 Mark the current recording with the target tag information.
  • the recording APP automatically uses the target tag information to mark the current recording.
  • the voiceprint can be used as the unique feature distinguishing the speaker because the speaker's voiceprint is unique. Segment a recording.
  • a voiceprint analysis is performed on the current recording, and different markings are used for marking different voiceprints.
  • the same speaker is tagged with the same tag. For example, a recording has 5 speakers, using the left slash, the right slash, the horizontal line, the vertical line, and the grid to mark the speakers A, B, C, D, and E, respectively. Among them, when speaker A has two speeches separated by other speakers in this recording, both speeches will use the left slash to mark speaker A to indicate the recording segment of the same speaker.
  • the speaker can be marked with a different color, for example, the speakers A, B, C, D, E are marked with red, yellow, blue, green and purple, respectively. Or when Speaker A has two speeches separated by other speakers in this recording, both of these speeches will be marked with a red marker to indicate that they are the recorded passage of the same speaker.
  • the recording mark method obtains a target voiceprint template of the voiceprint parameter by acquiring a current recording and extracting a voiceprint feature parameter from the current recording, and performing voiceprint clustering training on the voiceprint parameter. Determining whether the target voiceprint template is a voiceprint template in the voiceprint database, and if the determination result is yes, acquiring target marker information corresponding to the target voiceprint template from the voiceprint database, using the target marker The information marks the current recording.
  • the voiceprint template corresponding to the current recording is identified by the voiceprint, and the marker information corresponding to the current recording is obtained by using the established voiceprint database, thereby marking the current recording, thereby realizing the function of automatically marking the recording, and saving the user to add. The time of the tag.
  • FIG. 3 it is a schematic flowchart of a recording and marking method according to Embodiment 2 of the present application, and the recording marking method includes:
  • Step 201 Acquire a current recording and extract a voiceprint feature parameter from the current recording.
  • Step 202 Perform voiceprint clustering training on the voiceprint parameter to obtain a target voiceprint template of the voiceprint parameter.
  • Step 203 Determine whether the target voiceprint template is a voiceprint template in the voiceprint database.
  • step 204 If the result of the determination is yes, go to step 204; otherwise, go to step 205.
  • Step 204 Obtain target tag information corresponding to the target voiceprint template from the voiceprint database.
  • Step 205 Generate the target tag information corresponding to the target voiceprint template.
  • the recording APP may set a target marker information for the target voiceprint template to mark the target voiceprint template by the target marker information.
  • Step 206 Mark the current recording with the target tag information.
  • Step 207 Establish a mapping relationship between the target voiceprint template and the target tag information, and store the data in the voiceprint database.
  • Step 208 Receive remark information sent by the user through the terminal.
  • Step 209 Remark the current recording by using the remark information.
  • Step 210 Update the remark information to the target tag information in the voiceprint data.
  • the remark information may be the source name of the current recording.
  • the recording APP is instructed to use the remark information to remark the current recording.
  • the recording app can add a label to the location corresponding to the current recording.
  • the recording APP may further update the obtained remark information into the target mark information corresponding to the target voiceprint template corresponding to the current recording in the voiceprint data, The sound that can be recorded can be called again when it is the source corresponding to the current recording.
  • FIG. 4 it is a schematic diagram of an application example of the embodiment.
  • the recording APP automatically marks the current recording
  • the user can send a remark information to the recording APP through the terminal, which is used to add each speaker in the recording.
  • Remarks For example, the user can note the speaker A marked with a left slash as "Zhang Teacher" through the recording app.
  • FIG. 5 it is a schematic diagram of an application example of the embodiment.
  • the voice recording of the speaker will be directly after the voiceprint analysis. Mark as saved tag information.
  • the speaker A who has saved the previous recording is “Zhang Teacher”.
  • the new recording containing the speaker will not display the speaker A's mark, but the “Zhang Teacher”.
  • the recording includes the marking information corresponding to the speaker saved by the user, and the recorded recording needs to be quickly located according to the marked speaker. For example, if the user wants to find a recording of Mr. Zhang's lecture, just look for the label of "Zhang Teacher".
  • a voiceprint database needs to be created by the sample sound.
  • FIG. 7 it is a schematic flowchart of a method for establishing a voiceprint database in Embodiment 2 of the present application, and the method for establishing a voiceprint database includes:
  • Step 301 Analyze the sample sound, and extract the voiceprint feature parameter of the sample sound.
  • each recorded sound of the recording APP before the current recording is taken as the sample sound.
  • the recording APP analyzes the sampled sound of the recording, and extracts the voiceprint characteristic parameters of the sample sound, wherein the voiceprint characteristic parameters include: sound energy, formant, MFCC, and LPC.
  • Step 302 Perform voiceprint clustering training according to the voiceprint feature parameter of the sample sound to generate a sample voiceprint template.
  • the voiceprint clustering training is performed on the voiceprint feature parameters of the sample sound to generate the sample voiceprint template. If it is determined that the voiceprint characteristic parameters of the sample voiceprint have no similarity, the voiceprint feature parameters need to be buffered, and then the sound feature parameters are judged to have similarity, and then the voiceprint clustering training is performed on the voiceprint feature parameters. Generate a sample voiceprint template.
  • the 5 speakers can complete the sample sound. After training through voiceprint clustering, the 5 speakers can be identified as speakers A, B, and C. , D, E, and 5 speakers generate the corresponding sample voiceprint template.
  • Step 303 Generate corresponding sample tag information for the sample voiceprint template.
  • the speakers A, B, C, D, and E can be marked using a left oblique line, a right oblique line, a horizontal line, a vertical line, and a grid.
  • Step 304 Generate the voiceprint database by using the sample voiceprint template, the sample marker information, and a mapping relationship between the sample voiceprint template and the sample marker information.
  • the voiceprint database is generated by using a sample voiceprint template, the sample mark information, and a mapping relationship between the sample voiceprint template and the sample mark information.
  • the voiceprint template generated after each voiceprint clustering training of the recording is saved as a sample voiceprint template in the voiceprint database, and the mapping information of the sample voiceprint template and the mapping relationship between the two are also It will be saved to the voiceprint database to update the voiceprint database. In this way, when the recording of the same speaker is encountered again, the recording APP can quickly mark the speaker's recording through the voiceprint analysis, thereby improving the convenience of the recording mark.
  • the voiceprint template corresponding to the current recording is identified by the voiceprint, and the marked information corresponding to the current recording is obtained by using the established voiceprint database, thereby marking the current recording, thereby realizing the function of automatically marking the recording, and saving the user. The time when the tag was added.
  • FIG. 8 is a schematic structural diagram of a recording apparatus according to Embodiment 3 of the present application.
  • the apparatus includes an extraction module 11, a training module 12, a determination module 13, an acquisition module 14, and a marking module 15.
  • the extraction module 11 is configured to collect a current recording and extract a voiceprint feature parameter from the current recording.
  • the extraction module 11 performs feature analysis on the collected current recording, thereby obtaining the acoustic characteristic parameters of the current recording, wherein the acoustic characteristic parameters include: energy of the sound, formant MFCC, and LPC.
  • the training module 12 is configured to perform voiceprint clustering training on the voiceprint parameters to obtain a target voiceprint template of the voiceprint parameter.
  • a training module 12 in order to identify the template of the recording, a training module 12 is provided, wherein the training module 12 may be a voiceprint clustering trainer. After the voiceprint feature parameters are obtained, the voiceprint clustering training is performed on the voiceprint feature parameters by the training module 12, and the target voiceprint template corresponding to the current recording can be obtained.
  • the determining module 13 is configured to determine whether the target voiceprint template is a voiceprint template in the voiceprint database.
  • the sound pattern clustering training is performed on the sample sound by the training module 12, and the sample voiceprint template corresponding to the sample sound is obtained, and a voiceprint database is preset by using the sample voiceprint template.
  • a plurality of sample voiceprint templates are stored in the general voiceprint database, so that the user can perform recording marks during the recording process.
  • the determining module 13 may perform a search in the voiceprint database to determine whether the target voiceprint template exists in the voiceprint database.
  • the obtaining module 14 is configured to acquire target tag information corresponding to the target voiceprint template from the voiceprint database when the determination result of the determining module is YES.
  • the obtaining module 14 may acquire the target marker information corresponding to the target voiceprint template.
  • the marking module 15 is configured to mark the current recording with the target marking information.
  • the tagging module 15 After the target tag information is acquired, the tagging module 15 automatically uses the target tag information to tag the current recording.
  • the recording apparatus obtains the current voice recording and extracts the voiceprint feature parameters from the current recording, and performs voiceprint clustering training on the voiceprint parameters to obtain a target voiceprint template of the voiceprint parameter, and determines Whether the target voiceprint template is a voiceprint template in the voiceprint database, and if the determination result is yes, the target marker information corresponding to the target voiceprint template is obtained from the voiceprint database, and the target marker information is used. Mark the current recording.
  • the voiceprint template corresponding to the current recording is identified by the voiceprint, and the marker information corresponding to the current recording is obtained by using the established voiceprint database, thereby marking the current recording, thereby realizing the function of automatically marking the recording, and saving the user to add. The time of the tag.
  • FIG. 9 is a schematic structural diagram of a video recording apparatus according to Embodiment 4 of the present application.
  • the device includes the extraction module 11, the training module 12, the determination module 13, the acquisition module 14, and the marking module 15 in the third embodiment, and further includes a generating module 16, an establishing module 17, and a receiving module 18.
  • the generating module 16 is configured to generate the target tag information corresponding to the target voiceprint template when the determining result of the determining module is negative.
  • the establishing module 17 is configured to establish a mapping relationship between the target voiceprint template and the target tag information after the tag module 15 marks the current recording by using the target tag information, and store the sound in the sound In the database.
  • the extracting module 11 is further configured to analyze the sample sound to extract the voiceprint feature parameter of the sample sound before acquiring the current recording and extracting the voiceprint feature parameter from the current recording.
  • the training module 12 is further configured to generate a sample voiceprint template by performing voiceprint clustering training according to the voiceprint feature parameter of the sample sound.
  • the generating module 16 is further configured to generate a corresponding sample target for the sample voiceprint template. Remember the information.
  • the establishing module 17 is further configured to generate the voiceprint database by using the sample voiceprint template, the sample marker information, and a mapping relationship between the sample voiceprint template and the sample marker information.
  • the training module 12 is specifically configured to acquire the voiceprint feature parameter of the sample sound in a preset time period, and the voiceprint feature parameter of the sample sound in the preset time Similarly, the voiceprint clustering training is performed on the voiceprint feature parameter of the sample sound to generate the sample voiceprint template.
  • the receiving module 18 is configured to receive the remark information sent by the user through the terminal after the marking module 15 marks the current recording by using the target tag information;
  • the marking module 15 is further configured to use the remark information to remark the current recording;
  • the establishing module 17 is further configured to update the remark information into the target tag information in the voiceprint data.
  • the function modules of the recording apparatus provided in this embodiment can be used to execute the flow of the recording and marking method shown in FIG. 1, 3 and 7.
  • the specific working principle is not described here. For details, refer to the description of the method embodiment.
  • the recording apparatus obtains the current voice recording and extracts the voiceprint feature parameters from the current recording, and performs voiceprint clustering training on the voiceprint parameters to obtain a target voiceprint template of the voiceprint parameter, and determines Whether the target voiceprint template is a voiceprint template in the voiceprint database, and if the determination result is yes, the target marker information corresponding to the target voiceprint template is obtained from the voiceprint database, and the target marker information is used. Mark the current recording.
  • the voiceprint template corresponding to the current recording is identified by the voiceprint, and the marker information corresponding to the current recording is obtained by using the established voiceprint database, thereby marking the current recording, thereby realizing the function of automatically marking the recording, and saving the user to add. The time of the tag.
  • FIG. 10 is a schematic structural diagram of still another embodiment of a recording apparatus provided by the present application.
  • the recording apparatus of the embodiment of the present application includes a memory 61, one or more processors 62, and one or more programs 63.
  • the one or more programs 63 when executed by one or more processors 62, perform any of the above-described embodiments.
  • the recording device of the embodiment of the present application obtains the current voice recording and extracts the voiceprint feature parameters from the current recording, and performs voiceprint clustering training on the voiceprint parameters to obtain a target voiceprint template of the voiceprint parameter, and determines Whether the target voiceprint template is a voiceprint template in the voiceprint database, and if the determination result is yes, the target marker information corresponding to the target voiceprint template is obtained from the voiceprint database, and the target marker information is used. Mark the current recording.
  • the voiceprint template corresponding to the current recording is identified by the voiceprint, and the marker information corresponding to the current recording is obtained by using the established voiceprint database, thereby marking the current recording, thereby realizing the function of automatically marking the recording, and saving the user. The time when the tag was added.
  • FIG. 11 is a schematic structural diagram of an embodiment of a computer program product for recording marks provided by the present application.
  • the computer program product 71 for recording marks of the embodiment of the present application may include a signal bearing medium 72.
  • Signal bearing medium 72 may include one or more instructions 73 that, when executed by, for example, a processor, may provide the functionality described above with respect to Figures 1-9.
  • the instruction 73 may include: one or more instructions for acquiring a current recording and extracting a voiceprint feature parameter from the current recording; and performing voiceprint clustering training on the voiceprint parameter to obtain the voiceprint One or more instructions of the target voiceprint template of the parameter; one or more instructions for determining whether the target voiceprint template is a voiceprint template in the voiceprint database; for if the determination result is yes, from the One or more instructions for acquiring target tag information corresponding to the target voiceprint template in the voiceprint database; and one or more instructions for tagging the current recording using the target tag information.
  • the recording device can perform one or more of the steps shown in FIG. 1 in response to instruction 73.
  • signal bearing medium 72 can include computer readable media 74 such as, but not limited to, a hard disk drive, a compact disk (CD), a digital versatile disk (DVD), a digital tape, a memory, and the like.
  • the signal bearing medium 72 can include a recordable medium 75 such as, but not limited to, a memory, a read/write (R/W) CD, an R/W DVD, and the like.
  • the signal bearing medium 72 can include a communication medium 76 such as, but not limited to, a digital and/or analog communication medium (eg, fiber optic cable, waveguide, wired communication link, wireless communication link, etc.).
  • the computer program product 71 can be transmitted by the RF signal bearing medium 72 to one or more modules of the identification device of the multi-finger swipe gesture, wherein the signal bearing medium 72 is comprised of a wireless communication medium (eg, wireless compliant with the IEEE 802.11 standard) Communication medium) transmission.
  • a wireless communication medium eg, wireless compliant with the IEEE 802.11 standard
  • the computer program product of the embodiment of the present application obtains a voiceprint feature by acquiring a current recording and extracting a voiceprint feature parameter from the current recording, and obtaining a target voiceprint template of the voiceprint parameter by performing voiceprint clustering training on the voiceprint parameter. Determining whether the target voiceprint template is a voiceprint template in the voiceprint database, and if the determination result is yes, acquiring target marker information corresponding to the target voiceprint template from the voiceprint database, using the target marker The information marks the current recording.
  • the voiceprint template corresponding to the current recording is identified by the voiceprint, and the marker information corresponding to the current recording is obtained by using the established voiceprint database, thereby marking the current recording, thereby realizing the function of automatically marking the recording, and saving the user. The time when the tag was added.

Landscapes

  • Engineering & Computer Science (AREA)
  • Health & Medical Sciences (AREA)
  • Audiology, Speech & Language Pathology (AREA)
  • Human Computer Interaction (AREA)
  • Physics & Mathematics (AREA)
  • Acoustics & Sound (AREA)
  • Multimedia (AREA)
  • Computational Linguistics (AREA)
  • Artificial Intelligence (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Reverberation, Karaoke And Other Acoustics (AREA)
  • User Interface Of Digital Computer (AREA)

Abstract

录音标记方法及录音装置，通过采集当前录音并从所述当前录音中提取声纹特征参数（101），对所述声纹参数进行声纹聚类训练得到所述声纹参数的目标声纹模板（102），判断所述目标声纹模板是否为声纹数据库中的声纹模板（103），如果判断结果为是，从所述声纹数据库中获取与所述目标声纹模板对应的目标标记信息（104），使用所述目标标记信息对所述当前录音进行标记（105）。通过声纹识别当前录音对应的声纹模板，利用建立的声纹数据库获取与当前录音对应的标记信息，进而对当前录音进行标记，实现了自动标记录音的功能，而且节省了用户添加标记的时间。

Description

录音标记方法及录音装置
本专利申请要求申请日为2015年11月15日、申请号为2015107827293的中国专利申请的优先权，并将上述专利申请以引用的方式全文引入本文中。
技术领域
本申请实施例涉及电子技术领域，尤其涉及一种录音标记方法及录音装置。
背景技术
目前智能手机逐渐融入到了人们日常生活之中，不但成为日常通讯设备，也成为日常易于携带的记录设备。其中，用户通过智能手机的录音应用程序(Application，简称APP)可以对语音信息进行录制和保存，便于用户快速地保存一段难以直接记忆语音信息，并且还可以多次使用该录音。
现在很多录音APP可以让用户在录音的过程中，手动地对录音进行标记。通过这种手动地打标机的方式，便于整理和使用保存后的录音信息。当用户需要打标记时，只需要在录音时点击标设置好的记按钮，此时就会在录音的位置出现一个标记，在播放录音时该标记也会显示在播放条上。
在录制多人对话的过程中，当讲话者发生变化时，用户就需要相应地在录音中变更与当前讲话人对应的标记，这种通过手动打标记的方法便捷性较差，而且使得用户也无法关注录音的内容，需要花费用户的很多时间。
发明内容
本申请实施例提供一种录音标记方法及录音装置，用于解决手动标记时存在便捷性较差以及浪费时间的问题。
为了实现上述目的，本申请实施例提供了一种录音标记方法，包括：
采集当前录音并从所述当前录音中提取声纹特征参数；
对所述声纹参数进行声纹聚类训练得到所述声纹参数的目标声纹模板；
判断所述目标声纹模板是否为声纹数据库中的声纹模板；
如果判断结果为是，从所述声纹数据库中获取与所述目标声纹模板对应的目标标记信息；
使用所述目标标记信息对所述当前录音进行标记。
为了实现上述目的，本申请实施例提供了一种录音装置，包括：
提取模块，用于采集当前录音并从所述当前录音中提取声纹特征参数；
训练模块，用于对所述声纹参数进行声纹聚类训练得到所述声纹参数的目标声纹模板；
判断模块，用于判断所述目标声纹模板是否为声纹数据库中的声纹模板；
获取模块，用于在所述判断模块的判断结果为是时，从所述声纹数据库中获取与所述目标声纹模板对应的目标标记信息；
标记模块，用于使用所述目标标记信息对所述当前录音进行标记。
另一方面，本申请实施例提供一种录音装置，包括存储器、一个或多个处理器以及一个或多个程序，其中，所述一个或多个程序在由所述一个或多个处理器执行时执行下述操作：采集当前录音并从所述当前录音中提取声纹特征参数；对所述声纹参数进行声纹聚类训练得到所述声纹参数的目标声纹模板；判断所述目标声纹模板是否为声纹数据库中的声纹模板；如果判断结果为是，从所述声纹数据库中获取与所述目标声纹模板对应的目标标记信息；使用所述目标标记信息对所述当前录音进行标记。
另一方面，本申请实施例提供一种计算机可读存储介质，所述计算机可读存储介质上存储有计算机可执行指令，所述计算机可执行指令响应于执行使得录音装置执行操作，所述操作包括：采集当前录音并从所述当前录音中提取声纹特征参数；对所述声纹参数进行声纹聚 类训练得到所述声纹参数的目标声纹模板；判断所述目标声纹模板是否为声纹数据库中的声纹模板；如果判断结果为是，从所述声纹数据库中获取与所述目标声纹模板对应的目标标记信息；使用所述目标标记信息对所述当前录音进行标记。
本申请实施例的录音标记方法及录音装置，通过采集当前录音并从所述当前录音中提取声纹特征参数，对所述声纹参数进行声纹聚类训练得到所述声纹参数的目标声纹模板，判断所述目标声纹模板是否为声纹数据库中的声纹模板，如果判断结果为是，从所述声纹数据库中获取与所述目标声纹模板对应的目标标记信息，使用所述目标标记信息对所述当前录音进行标记。本申请实施例通过声纹识别当前录音对应的声纹模板，利用建立的声纹数据库获取与当前录音对应的标记信息，进而对当前录音进行标记，实现了自动标记录音的功能，而且节省了用户添加标记的时间。
附图说明
图1为本申请实施例一的录音标记方法的流程示意图；
图2为本申请实施例一的录音标记方法的应用示例示意图；
图3为本申请实施例二的录音标记方法的流程示意图；
图4为本申请实施例二的录音标记方法的应用示例示意图之一；
图5为本申请实施例一的录音标记方法的应用示例示意图之二；
图6为本申请实施例一的录音标记方法的应用示例示意图之三；
图7为本申请实施例二中的声纹数据库建立方法的流程示意图；
图8为本申请实施例三的录音装置的结构示意图；
图9为本申请实施例四的录音装置的结构示意图；
图10为本申请提供的录音装置又一个实施例的结构示意图；
图11为本申请提供的用于录音标记的计算机程序产品一个实施例的结构示意图。
具体实施方式
下面结合附图对本申请实施例提供的录音标记方法及录音装置进行详细描述。
实施例一
如图1所示，其为本申请实施例一的录音标记方法的流程示意图，该录音标记方法包括：
步骤101、采集当前录音并从所述当前录音中提取声纹特征参数。
用户可以通过智能手机的用户界面，开启智能手机中下载的录音APP的录音功能，录音APP开始对当前录音进行采集，在采集的过程中录音APP可以对声音进行预处理，例如，采集的数据进行分帧、加窗和滤波等。
进一步地，对采集的当前录音进行特征分析，进而得到当前录音的声波特征参数，其中，声波特征参数包括：声音的能量、共振峰、梅尔倒谱系数(Mel-frequency cepstrum coefficients，简称MFCC)以及线性预测系数(Linear Prediction Coefficients，简称LPC)。
步骤102、对所述声纹参数进行声纹聚类训练得到所述声纹参数的目标声纹模板。
本实施例中，为了识别出录音的模板，设置有声纹聚类训练器，在获取到声纹特征参数后，通过该训练器对声纹特征参数进行声纹聚类训练，就可以得到该当前录音对应的目标声纹模板。
步骤103、判断所述目标声纹模板是否为声纹数据库中的声纹模板。
本实施例中，通过训练器对样本声音进行声纹聚类训练，得到了样本声音对应的样本声纹模板，使用样本声纹模板预先设置了一个声纹数据库存储在录音APP中。一般声纹数据库中存储有多个样本声纹模板，以便于用户在录音过程中进行录音标记。在获取到目标声纹模板后，录音APP可以在声纹数据库中进行查找，判断该目标声纹模板是否存在于该声纹数据库中。
步骤104、若目标声纹模板为声纹数据库中的声纹模板，从所述声纹数据库中获取与所述目标声纹模板对应的目标标记信息。
在声纹数据库中不仅保存有样本声纹模板，而且还存储有与样本声纹模板对应的标记信息，一般每个样本声纹模板对应有各自的标记信息。当在声纹数据库中获取到与目标声纹模板对应的样本声纹模板时，就可以获取与该目标声纹模板对应的目标标记信息。
步骤105、使用所述目标标记信息对所述当前录音进行标记。
在获取到目标标记信息后，录音APP自动使用该目标标记信息对当前录音进行标记。
如图2所示，其为本实施例的应用示例示意图，由于说话人的声纹的唯一性可以利用声纹作为区别说话人的唯一特征。对一段录音进行分段标记。首先对当前录音进行声纹分析，对于不同声纹使用不同的标记进行标记。同一个说话人使用相同的标记进行标记。比如一段录音有5个说话人，分别使用左斜线、右斜线、横线、竖线以及网格进行标记说话人A、B、C、D、E。其中，当说话人A在这段录音中有两次被其他说话人分开的发言，这两段发言都会使用左斜线标记说话人A，以表明是同一个说话人的录音段落。为了用户更直观的看到说话人的不同，可以使用不同的颜色标记说话人，例如，分别使用红色、黄色、蓝色、绿色和紫色来标记说话人A、B、C、D、E。或者当说话人A在这段录音中有两次被其他说话人分开的发言，这两段发言都会使用红色标记说话人A，以表明是同一个说话人的录音段落。
本实施例提供的录音标记方法，通过采集当前录音并从所述当前录音中提取声纹特征参数，对所述声纹参数进行声纹聚类训练得到所述声纹参数的目标声纹模板，判断所述目标声纹模板是否为声纹数据库中的声纹模板，如果判断结果为是，从所述声纹数据库中获取与所述目标声纹模板对应的目标标记信息，使用所述目标标记信息对所述当前录音进行标记。本实施例通过声纹识别当前录音对应的声纹模板，利用建立的声纹数据库获取与当前录音对应的标记信息，进而对当前录音进行标记，实现了自动标记录音的功能，而且节省了用户添加标记的时间。
实施例二
如图3所示，其为本申请实施例二的录音标记方法的流程示意图，该录音标记方法包括：
步骤201、采集当前录音并从所述当前录音中提取声纹特征参数。
步骤202、对所述声纹参数进行声纹聚类训练得到所述声纹参数的目标声纹模板。
步骤203、判断所述目标声纹模板是否为声纹数据库中的声纹模板
如果判断结果为是，执行步骤204；否则执行步骤205。
步骤204、从所述声纹数据库中获取与所述目标声纹模板对应的目标标记信息。
步骤201～步骤204的相关内容可参见上述实施例一中步骤101～步骤104中的记载，此处不再赘述。
步骤205、生成与所述目标声纹模板对应的所述目标标记信息。
在识别出目标声纹模板并不存在与声纹数据库中之后，录音APP可以为该目标声纹模板设置一个目标标记信息，以通过该目标标记信息对该目标声纹模板进行标记。
步骤206、使用所述目标标记信息对所述当前录音进行标记。
步骤207、建立所述目标声纹模板与所述目标标记信息之间映射关系并存储在所述声纹数据库中。
步骤208、接收用户通过终端发送的备注信息。
步骤209、使用所述备注信息对所述当前录音进行备注。
步骤210、将所述备注信息更新到所述声纹数据中所述目标标记信息中。
接收用户通过终端发送的备注信息，备注信息可以为当前录音的来源名称，在终端获取到备注信息后，指示录音APP使用该备注信息对当前录音进行备注。例如，录音APP可以为当前录音对应的位置添加一个标签。进一步地，录音APP还可以将获取到的备注信息更新到声纹数据中与当前录音对应的目标声纹模板对应的目标标记信息中， 以便录制的声音为当前录音对应的音源时可以再次被调用。
如图4所示，其为本实施例的应用示例示意图，当录音APP对当前录音进行自动标记后，用户可以通过终端向录音APP发送备注信息，用于给这段录音中每位说话人添加备注信息。比如，用户可以通过录音APP将用左斜线标记的说话人A备注为“张老师”。用户可为新说话人添加的备注信息，并直接与该说话人的声纹信息匹配，并作为这段录音的名称。
如图5所示，其为本实施例的应用示例示意图，当用户新建一段录音，如果其中包含已保存声音名称的说话人的录音，在声纹分析后，这位说话人的录音段落会直接标记为已保存的标记信息。比如已经保存了之前一段录音的说话人A为“张老师”，新建一段包含这个说话人的录音不会再显示说话人A的标记，而是显示“张老师”。
如图6所示，其为本实施例的应用示例示意图，录音中包含用户保存过的讲话人对应的标记信息，按照所标记的说话人，更快定位需要寻找的录音。比如用户想要寻找张老师的讲课录音，只要寻找“张老师”的标签即可。
在步骤201采集当前录音并从所述当前录音中提取声纹特征参数之前，还需要通过样本声音建立一个声纹数据库。
如图7所示，其为本申请实施例二中的声纹数据库建立方法的流程示意图，该声纹数据库建立方法包括：
步骤301、对样本声音进行分析，提取所述样本声音的所述声纹特征参数。
本实施例中，将录音APP在当前录音之前的每次录制的声音作为样本声音。在获取到每次录音后，录音APP会对录音的样本声音进行分析，提取出该样本声音的声纹特征参数，其中声纹特征参数包括：声音的能量、共振峰、MFCC以及LPC等。
步骤302、根据所述样本声音的所述声纹特征参数进行声纹聚类训练生成样本声纹模板。
为了对获取到样本声纹的声纹特征参数进行声纹聚类训练，需要 进一步确定该声纹特征参数是否为同一个音源的声音，具体地，预设时间段内的所述样本声音的所述声纹特征参数，当所述预设时间内的所述样本声音的所述声纹特征参数具有相似性时，对所述样本声音的所述声纹特征参数进行声纹聚类训练生成所述样本声纹模板。如果确定出样本声纹的声纹特征参数的不具有相似性，则需要将声纹特征参数进行缓存，再判断出该声音特征参数具有相似性之后，对声纹特征参数进行声纹聚类训练生成样本声纹模板。
比如，有一段录音中有5个说话人，这5个说话人就可以做完样本声音，在通过声纹聚类训练后，可以识别出这个5个说话人分别为说话人A、B、C、D、E，并5个说话人生成相应的样本声纹模板。
步骤303、为所述样本声纹模板生成对应的样本标记信息。
在生成样本声纹模板后，为样本声音生成对应的样本标记信息，例如同一个说话人使用相同的标记进行标记。本实施例中，可以使用左斜线、右斜线、横线、竖线以及网格进行标记说话人A、B、C、D、E。
步骤304、使用所述样本声纹模板、所述样本标记信息以及所述样本声纹模板与所述样本标记信息之间的映射关系生成所述声纹数据库。
为了提高对录音标记的快捷性，本实施例中，使用样本声纹模板、所述样本标记信息以及所述样本声纹模板与所述样本标记信息之间的映射关系生成所述声纹数据库。每次对录音进行声纹聚类训练后生成的声纹模板都会作为样本声纹模板保存到声纹数据库中，而且会将对该样本声纹模板的标记信息以及两者之间的映射关系也会保存到声纹数据库中，以对声纹数据库进行更新。这样当再次遇到同一说话人的录音时，录音APP通过声纹分析，能够很迅速地对该说话人的录音进行标记，提高了录音标记的便捷性。
本实施例通过声纹识别当前录音对应的声纹模板，并利用建立的声纹数据库获取与当前录音对应的标记信息，进而对当前录音进行标记，实现了自动标记录音的功能，而且节省了用户添加标记的时间。
实施例三
如图8所示，其为本申请实施例三的录音装置的结构示意图。该装置包括：提取模块11、训练模块12、判断模块13、获取模块14和标记模块15。
其中，提取模块11，用于采集当前录音并从所述当前录音中提取声纹特征参数。
提取模块11对采集的当前录音进行特征分析，进而得到当前录音的声波特征参数，其中，声波特征参数包括：声音的能量、共振峰MFCC以及LPC。
训练模块12，用于对所述声纹参数进行声纹聚类训练得到所述声纹参数的目标声纹模板。
本实施例中，为了识别出录音的模板，设置有训练模块12，其中训练模块12可以为声纹聚类训练器。在获取到声纹特征参数后，通过该训练模块12对声纹特征参数进行声纹聚类训练，就可以得到该当前录音对应的目标声纹模板。
判断模块13，用于判断所述目标声纹模板是否为声纹数据库中的声纹模板。
本实施例中，通过训练模块12对样本声音进行声纹聚类训练，得到了样本声音对应的样本声纹模板，使用样本声纹模板预先设置了一个声纹数据库。一般声纹数据库中存储有多个样本声纹模板，以便于用户在录音过程中进行录音标记。在获取到目标声纹模板后，判断模块13可以在声纹数据库中进行查找，判断该目标声纹模板是否存在于该声纹数据库中。
获取模块14，用于在所述判断模块的判断结果为是时，从所述声纹数据库中获取与所述目标声纹模板对应的目标标记信息。
在声纹数据库中不仅保存有样本声纹模板，而且还存储有与样本声纹模板对应的标记信息，一般每个样本声纹模板对应有各自的标记信息。当在声纹数据库中获取到与目标声纹模板对应的样本声纹模板时，获取模块14就可以获取与该目标声纹模板对应的目标标记信息。
标记模块15，用于使用所述目标标记信息对所述当前录音进行标记。
在获取到目标标记信息后，标记模块15自动使用该目标标记信息对当前录音进行标记。
本实施例提供的录音装置，通过采集当前录音并从所述当前录音中提取声纹特征参数，对所述声纹参数进行声纹聚类训练得到所述声纹参数的目标声纹模板，判断所述目标声纹模板是否为声纹数据库中的声纹模板，如果判断结果为是，从所述声纹数据库中获取与所述目标声纹模板对应的目标标记信息，使用所述目标标记信息对所述当前录音进行标记。本实施例通过声纹识别当前录音对应的声纹模板，利用建立的声纹数据库获取与当前录音对应的标记信息，进而对当前录音进行标记，实现了自动标记录音的功能，而且节省了用户添加标记的时间。
实施例四
图9所示，其为本申请实施例四的录像装置的结构示意图。该装置包括上述实施例三中的提取模块11、训练模块12、判断模块13、获取模块14和标记模块15之外，还包括生成模块16、建立模块17和接收模块18。
其中，所述生成模块16，用于在所述判断模块的判断结果为否时，生成与所述目标声纹模板对应的所述目标标记信息。
所述建立模块17，用于在标记模块15使用所述目标标记信息对所述当前录音进行标记之后，建立所述目标声纹模板与所述目标标记信息之间映射关系并存储在所述声纹数据库中。
进一步地，所述提取模块11，还用于在采集当前录音并从所述当前录音中提取声纹特征参数之前，对样本声音进行分析提取所述样本声音的所述声纹特征参数。
所述训练模块12，还用于根据所述样本声音的所述声纹特征参数进行声纹聚类训练生成样本声纹模板。
所述生成模块16，还用于为所述样本声纹模板生成对应的样本标 记信息。
所述建立模块17，还用于使用所述样本声纹模板、所述样本标记信息以及所述样本声纹模板与所述样本标记信息之间的映射关系生成所述声纹数据库。
进一步地，所述训练模块12，具体用于获取预设时间段内的所述样本声音的所述声纹特征参数，在所述预设时间内的所述样本声音的所述声纹特征参数相似时，对所述样本声音的所述声纹特征参数进行声纹聚类训练生成所述样本声纹模板。
进一步地，接收模块18，用于在所述标记模块15使用所述目标标记信息对所述当前录音进行标记之后，接收用户通过终端发送的备注信息；
所述标记模块15，还用于使用所述备注信息对所述当前录音进行备注；
所述建立模块17，还用于将所述备注信息更新到所述声纹数据中所述目标标记信息中。
本实施例提供的录音装置的各功能模块可用于执行图1、3和7所示的录音标记方法的流程，其具体工作原理不再赘述，详见方法实施例的描述。
本实施例提供的录音装置，通过采集当前录音并从所述当前录音中提取声纹特征参数，对所述声纹参数进行声纹聚类训练得到所述声纹参数的目标声纹模板，判断所述目标声纹模板是否为声纹数据库中的声纹模板，如果判断结果为是，从所述声纹数据库中获取与所述目标声纹模板对应的目标标记信息，使用所述目标标记信息对所述当前录音进行标记。本实施例通过声纹识别当前录音对应的声纹模板，利用建立的声纹数据库获取与当前录音对应的标记信息，进而对当前录音进行标记，实现了自动标记录音的功能，而且节省了用户添加标记的时间。
实施例五
图10为本申请提供的录音装置又一个实施例的结构示意图。如图10所示，本申请实施例的录音装置包括：存储器61、一个或多个处理器62以及一个或多个程序63。
其中，所述一个或多个程序63在由一个或多个处理器62执行时执行上述实施例中的任意一种方法。
本申请实施例的录音装置，通过采集当前录音并从所述当前录音中提取声纹特征参数，对所述声纹参数进行声纹聚类训练得到所述声纹参数的目标声纹模板，判断所述目标声纹模板是否为声纹数据库中的声纹模板，如果判断结果为是，从所述声纹数据库中获取与所述目标声纹模板对应的目标标记信息，使用所述目标标记信息对所述当前录音进行标记。本申请实施例通过声纹识别当前录音对应的声纹模板，利用建立的声纹数据库获取与当前录音对应的标记信息，进而对当前录音进行标记，实现了自动标记录音的功能，而且节省了用户添加标记的时间。
实施例六
图11为本申请提供的用于录音标记的计算机程序产品一个实施例的结构示意图。如图11所示，本申请实施例的用于录音标记的计算机程序产品71，可以包括信号承载介质72。信号承载介质72可以包括一个或更多个指令73，该指令73在由例如处理器执行时，处理器可以提供以上针对图1-9描述的功能。例如，指令73可以包括：用于采集当前录音并从所述当前录音中提取声纹特征参数的一个或多个指令；用于对所述声纹参数进行声纹聚类训练得到所述声纹参数的目标声纹模板的一个或多个指令；用于判断所述目标声纹模板是否为声纹数据库中的声纹模板的一个或多个指令；用于如果判断结果为是，从所述声纹数据库中获取与所述目标声纹模板对应的目标标记信息的一个或多个指令；以及用于使用所述目标标记信息对所述当前录音进行标记的一个或多个指令。因此，例如，参照图8，录音装置可以响应于指令73来进行图1中所示的步骤中的一个或更多个。
在一些实现中，信号承载介质72可以包括计算机可读介质74，诸如但不限于硬盘驱动器、压缩盘(CD)、数字通用盘(DVD)、数字带、存储器等。在一些实现中，信号承载介质72可以包括可记录介质75，诸如但不限于存储器、读/写(R/W)CD、R/W DVD等。在一些实现中，信号承载介质72可以包括通信介质76，诸如但不限于数字和/或模拟通信介质(例如，光纤线缆、波导、有线通信链路、无线通信链路等)。因此，例如，计算机程序产品71可以通过RF信号承载介质72传送给多指滑动手势的识别装置的一个或多个模块，其中，信号承载介质72由无线通信介质(例如，符合IEEE 802.11标准的无线通信介质)传送。
本申请实施例的计算机程序产品，通过采集当前录音并从所述当前录音中提取声纹特征参数，对所述声纹参数进行声纹聚类训练得到所述声纹参数的目标声纹模板，判断所述目标声纹模板是否为声纹数据库中的声纹模板，如果判断结果为是，从所述声纹数据库中获取与所述目标声纹模板对应的目标标记信息，使用所述目标标记信息对所述当前录音进行标记。本申请实施例通过声纹识别当前录音对应的声纹模板，利用建立的声纹数据库获取与当前录音对应的标记信息，进而对当前录音进行标记，实现了自动标记录音的功能，而且节省了用户添加标记的时间。
通过以上的实施方式的描述，本领域的技术人员可以清楚地了解到各实施方式可借助软件加必需的通用硬件平台的方式来实现，当然也可以通过硬件。基于这样的理解，上述技术方案本质上或者说对现有技术做出贡献的部分可以以软件产品的形式体现出来，该计算机软件产品可以存储在计算机可读存储介质中，如ROM/RAM、磁碟、光盘等，包括若干指令用以使得一台计算机设备(可以是个人计算机，服务器，或者网络设备等)执行各个实施例或者实施例的某些部分所述的方法。
最后应说明的是：以上各实施例仅用以说明本申请的技术方案，而非对其限制；尽管参照前述各实施例对本申请进行了详细的说明， 本领域的普通技术人员应当理解：其依然可以对前述各实施例所记载的技术方案进行修改，或者对其中部分或者全部技术特征进行等同替换；而这些修改或者替换，并不使相应技术方案的本质脱离本申请各实施例技术方案的范围。

Claims (12)

  1. 一种录音标记方法，其特征在于，包括：
    采集当前录音并从所述当前录音中提取声纹特征参数；
    对所述声纹参数进行声纹聚类训练得到所述声纹参数的目标声纹模板；
    判断所述目标声纹模板是否为声纹数据库中的声纹模板；
    如果判断结果为是，从所述声纹数据库中获取与所述目标声纹模板对应的目标标记信息；
    使用所述目标标记信息对所述当前录音进行标记。
  2. 根据权利要求1所述的录音标记方法，其特征在于，所述使用所述目标标记信息对所述当前录音进行标记之前，还包括：
    如果判断结果为否，生成与所述目标声纹模板对应的所述目标标记信息。
  3. 根据权利要求2所述的录音标记方法，其特征在于，所述使用所述目标标记信息对所述当前录音进行标记之后，还包括：
    建立所述目标声纹模板与所述目标标记信息之间映射关系并存储在所述声纹数据库中。
  4. 根据权利要求1-3任一项所述的录音标记方法，其特征在于，所述采集当前录音并从所述当前录音中提取声纹特征参数之前，包括：
    对样本声音进行分析，提取所述样本声音的所述声纹特征参数；
    根据所述样本声音的所述声纹特征参数进行声纹聚类训练生成样本声纹模板；
    为所述样本声纹模板生成对应的样本标记信息；
    使用所述样本声纹模板、所述样本标记信息以及所述样本声纹模板与所述样本标记信息之间的映射关系生成所述声纹数据库。
  5. 根据权利要求4所述的录音标记方法，其特征在于，所述根据所述样本声音的所述声纹特征参数进行声纹聚类训练生成样本声纹模板包括：
    获取预设时间段内的所述样本声音的所述声纹特征参数；
    在所述预设时间内的所述样本声音的所述声纹特征参数具有相似性时，对所述样本声音的所述声纹特征参数进行声纹聚类训练生成所述样本声纹模板。
  6. 根据权利要求1-3任一项所述的录音标记方法，其特征在于，所述使用所述目标标记信息对所述当前录音进行标记之后，还包括：
    接收用户通过终端发送的备注信息；
    使用所述备注信息对所述当前录音进行备注；
    将所述备注信息更新到所述声纹数据中所述目标标记信息中。
  7. 一种录音装置，其特征在于，包括：
    提取模块，用于采集当前录音并从所述当前录音中提取声纹特征参数；
    训练模块，用于对所述声纹参数进行声纹聚类训练得到所述声纹参数的目标声纹模板；
    判断模块，用于判断所述目标声纹模板是否为声纹数据库中的声纹模板；
    获取模块，用于在所述判断模块的判断结果为是时，从所述声纹数据库中获取与所述目标声纹模板对应的目标标记信息；
    标记模块，用于使用所述目标标记信息对所述当前录音进行标记。
  8. 根据权利要求7所述的录音装置，其特征在于，还包括：
    生成模块，用于在所述判断模块的判断结果为否时，生成与所述目标声纹模板对应的所述目标标记信息。
  9. 根据权利要求8所述的录音装置，其特征在于，还包括；
    建立模块，用于在所述标记模块使用所述目标标记信息对所述当前录音进行标记之后，建立所述目标声纹模板与所述目标标记信息之间映射关系并存储在所述声纹数据库中。
  10. 根据权利要求7-9任一项所述的录音装置，其特征在于，所述提取模块，还用于在采集当前录音并从所述当前录音中提取声纹特征参数之前，对样本声音进行分析提取所述样本声音的所述声纹特征参数；
    所述训练模块，还用于根据所述样本声音的所述声纹特征参数进行声纹聚类训练生成样本声纹模板；
    所述生成模块，还用于为所述样本声纹模板生成对应的样本标记信息；
    所述建立模块，还用于使用所述样本声纹模板、所述样本标记信息以及所述样本声纹模板与所述样本标记信息之间的映射关系生成所述声纹数据库。
  11. 根据权利要求10所述的录音装置，其特征在于，所述训练模块，具体用于获取预设时间段内的所述样本声音的所述声纹特征参数，在所述预设时间内的所述样本声音的所述声纹特征参数具有相似时，对所述样本声音的所述声纹特征参数进行声纹聚类训练生成所述样本声纹模板。
  12. 根据权利要求7-9任一项所述的录音装置，其特征在于，还包括：
    接收模块，用于在所述标记模块使用所述目标标记信息对所述当前录音进行标记之后，接收用户通过终端发送的备注信息；
    所述标记模块，还用于使用所述备注信息对所述当前录音进行备注；
    所述建立模块，还用于将所述备注信息更新到所述声纹数据中所述目标标记信息中。
PCT/CN2016/089033 2015-11-15 2016-07-07 录音标记方法及录音装置 Ceased WO2017080239A1 (zh)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN201510782729.3A CN105895077A (zh) 2015-11-15 2015-11-15 录音标记方法及录音装置
CN201510782729.3 2015-11-15

Publications (1)

Publication Number Publication Date
WO2017080239A1 true WO2017080239A1 (zh) 2017-05-18

Family

ID=57001881

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2016/089033 Ceased WO2017080239A1 (zh) 2015-11-15 2016-07-07 录音标记方法及录音装置

Country Status (2)

Country Link
CN (1) CN105895077A (zh)
WO (1) WO2017080239A1 (zh)

Cited By (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN111243601A (zh) * 2019-12-31 2020-06-05 北京捷通华声科技股份有限公司 声纹聚类方法、装置、电子设备和计算机可读存储介质
CN111292733A (zh) * 2018-12-06 2020-06-16 阿里巴巴集团控股有限公司 一种语音交互方法和装置
CN116206613A (zh) * 2023-05-06 2023-06-02 北京惠朗时代科技有限公司 基于智能家居场景的声纹识别管理系统

Families Citing this family (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN107452385A (zh) * 2017-08-16 2017-12-08 北京世纪好未来教育科技有限公司 一种基于语音的数据评价方法及装置
CN107564531A (zh) * 2017-08-25 2018-01-09 百度在线网络技术（北京）有限公司 基于声纹特征的会议记录方法、装置及计算机设备
CN107863108B (zh) * 2017-11-16 2021-03-23 百度在线网络技术（北京）有限公司 信息输出方法和装置
CN108109633A (zh) * 2017-12-20 2018-06-01 北京声智科技有限公司 无人值守的云端语音库采集与智能产品测试的系统与方法
WO2019205119A1 (zh) * 2018-04-28 2019-10-31 海能达通信股份有限公司 一种语音播放方法、装置及客户端
CN113763962A (zh) * 2021-05-10 2021-12-07 腾讯科技（深圳）有限公司 音频处理方法、装置、存储介质及计算机设备
CN115331673B (zh) * 2022-10-14 2023-01-03 北京师范大学 一种复杂声音场景下的声纹识别家电控制方法和装置

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2010060850A (ja) * 2008-09-04 2010-03-18 Nec Corp 議事録作成支援装置、議事録作成支援方法、議事録作成支援用プログラム及び議事録作成支援システム
CN102985965A (zh) * 2010-05-24 2013-03-20 微软公司 声纹标识
CN104427292A (zh) * 2013-08-22 2015-03-18 中兴通讯股份有限公司 会议纪要的提取方法及装置
CN104731927A (zh) * 2015-03-27 2015-06-24 努比亚技术有限公司 录音文件分类方法和系统
CN104978507A (zh) * 2014-04-14 2015-10-14 中国石油化工集团公司 一种基于声纹识别的智能测井评价专家系统身份认证方法

Family Cites Families (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN102760434A (zh) * 2012-07-09 2012-10-31 华为终端有限公司 一种声纹特征模型更新方法及终端
CN204242252U (zh) * 2014-11-18 2015-04-01 科大讯飞股份有限公司 一种带声纹识别功能的行车记录仪
CN104575492A (zh) * 2014-12-31 2015-04-29 深圳市航盛电子股份有限公司 一种声纹识别方法及装置和无钥匙车锁系统及实现方法

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2010060850A (ja) * 2008-09-04 2010-03-18 Nec Corp 議事録作成支援装置、議事録作成支援方法、議事録作成支援用プログラム及び議事録作成支援システム
CN102985965A (zh) * 2010-05-24 2013-03-20 微软公司 声纹标识
CN104427292A (zh) * 2013-08-22 2015-03-18 中兴通讯股份有限公司 会议纪要的提取方法及装置
CN104978507A (zh) * 2014-04-14 2015-10-14 中国石油化工集团公司 一种基于声纹识别的智能测井评价专家系统身份认证方法
CN104731927A (zh) * 2015-03-27 2015-06-24 努比亚技术有限公司 录音文件分类方法和系统

Cited By (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN111292733A (zh) * 2018-12-06 2020-06-16 阿里巴巴集团控股有限公司 一种语音交互方法和装置
CN111243601A (zh) * 2019-12-31 2020-06-05 北京捷通华声科技股份有限公司 声纹聚类方法、装置、电子设备和计算机可读存储介质
CN111243601B (zh) * 2019-12-31 2023-04-07 北京捷通华声科技股份有限公司 声纹聚类方法、装置、电子设备和计算机可读存储介质
CN116206613A (zh) * 2023-05-06 2023-06-02 北京惠朗时代科技有限公司 基于智能家居场景的声纹识别管理系统

Also Published As

Publication number Publication date
CN105895077A (zh) 2016-08-24

Similar Documents

Publication Publication Date Title
WO2017080239A1 (zh) 录音标记方法及录音装置
WO2017080235A1 (zh) 录音编辑方法及录音装置
CN105808782B (zh) 一种图片标签的添加方法及装置
CN107274916B (zh) 基于声纹信息对音频/视频文件进行操作的方法及装置
JP6688340B2 (ja) 表情アイコンを入力するための方法及び装置
JP4600828B2 (ja) 文書対応付け装置、および文書対応付け方法
CN1333363C (zh) 音频信号处理设备和音频信号处理方法
TW202008349A (zh) 語音標註方法、裝置及設備
CN107562760B (zh) 一种语音数据处理方法及装置
WO2019148586A1 (zh) 多人发言中发言人识别方法以及装置
CN106971723A (zh) 语音处理方法和装置、用于语音处理的装置
CN102256030A (zh) 可匹配背景音乐的相册演示系统及其背景音乐匹配方法
CN103530432A (zh) 一种具有语音提取功能的会议记录器及语音提取方法
CN108305618B (zh) 语音获取及搜索方法、智能笔、搜索终端及存储介质
WO2016197708A1 (zh) 一种录音方法及终端
CN113691382A (zh) 会议记录方法、装置、计算机设备及介质
CN106548793A (zh) 存储和播放音频文件的方法和装置
CN109947971B (zh) 图像检索方法、装置、电子设备及存储介质
CN106297841A (zh) 一种音频跟读引导方法及装置
WO2023160515A1 (zh) 视频处理方法、装置、设备及介质
CN110110147A (zh) 一种视频检索的方法及装置
CN109213970B (zh) 笔录生成方法及装置
Chen et al. A proof-of-concept study for automatic speech recognition to transcribe AAC speakers’ speech from high-technology AAC systems
CN109635151A (zh) 建立音频检索索引的方法、装置及计算机设备
CN115985315A (zh) 说话人标注方法、装置、电子设备和存储介质

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 16863417

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 16863417

Country of ref document: EP

Kind code of ref document: A1