WO2017084592A1 - 恶意电话的语音特征库的配置方法和装置 - Google Patents

恶意电话的语音特征库的配置方法和装置 Download PDF

Info

Publication number
WO2017084592A1
WO2017084592A1 PCT/CN2016/106150 CN2016106150W WO2017084592A1 WO 2017084592 A1 WO2017084592 A1 WO 2017084592A1 CN 2016106150 W CN2016106150 W CN 2016106150W WO 2017084592 A1 WO2017084592 A1 WO 2017084592A1
Authority
WO
WIPO (PCT)
Prior art keywords
phone
voice
malicious
malicious phone
voice feature
Prior art date
Application number
PCT/CN2016/106150
Other languages
English (en)
French (fr)
Inventor
王务志
王军
Original Assignee
北京奇虎科技有限公司
奇智软件(北京)有限公司
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 北京奇虎科技有限公司, 奇智软件(北京)有限公司 filed Critical 北京奇虎科技有限公司
Publication of WO2017084592A1 publication Critical patent/WO2017084592A1/zh

Links

Images

Classifications

    • GPHYSICS
    • G10MUSICAL INSTRUMENTS; ACOUSTICS
    • G10LSPEECH ANALYSIS TECHNIQUES OR SPEECH SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING TECHNIQUES; SPEECH OR AUDIO CODING OR DECODING
    • G10L17/00Speaker identification or verification techniques
    • G10L17/04Training, enrolment or model building
    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04MTELEPHONIC COMMUNICATION
    • H04M1/00Substation equipment, e.g. for use by subscribers
    • H04M1/66Substation equipment, e.g. for use by subscribers with means for preventing unauthorised or fraudulent calling
    • H04M1/663Preventing unauthorised calls to a telephone set
    • GPHYSICS
    • G10MUSICAL INSTRUMENTS; ACOUSTICS
    • G10LSPEECH ANALYSIS TECHNIQUES OR SPEECH SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING TECHNIQUES; SPEECH OR AUDIO CODING OR DECODING
    • G10L15/00Speech recognition
    • G10L15/02Feature extraction for speech recognition; Selection of recognition unit
    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04MTELEPHONIC COMMUNICATION
    • H04M1/00Substation equipment, e.g. for use by subscribers
    • H04M1/66Substation equipment, e.g. for use by subscribers with means for preventing unauthorised or fraudulent calling
    • H04M1/663Preventing unauthorised calls to a telephone set
    • H04M1/665Preventing unauthorised calls to a telephone set by checking the validity of a code
    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04MTELEPHONIC COMMUNICATION
    • H04M3/00Automatic or semi-automatic exchanges
    • H04M3/22Arrangements for supervision, monitoring or testing
    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04MTELEPHONIC COMMUNICATION
    • H04M3/00Automatic or semi-automatic exchanges
    • H04M3/42Systems providing special services or facilities to subscribers
    • H04M3/436Arrangements for screening incoming calls, i.e. evaluating the characteristics of a call before deciding whether to answer it

Definitions

  • the invention belongs to the technical field of mobile devices, and in particular to a method and a device for configuring a voice feature library of a malicious phone.
  • the present application provides a method and apparatus for configuring a voice feature library of a malicious phone, which can implement interception of a malicious phone by matching of voice feature points.
  • the embodiment of the invention provides a method for configuring a voice feature database of a malicious phone, which is executed on the user equipment side, and includes:
  • the other party's incoming call is identified as a malicious call, and the voice of the malicious phone is recorded for a certain period of time;
  • the voice performs voice analysis, obtains a voice feature point of the malicious phone, and saves the voice feature point of the malicious phone to the malicious phone voice feature database.
  • the malicious phone comprises an intermediary phone, a sales phone, a fraud phone, a harassing phone or a custom phone;
  • the other party's call is identified as a malicious call, including:
  • a phone label option is displayed on the call interface, including the label of the intermediary phone, the tag of the promotional phone, the tag of the scam phone, the tag of the harassing phone, or the option of the custom phone tag.
  • the user selects any label according to the voice of the other party in the phone label option, the user identifies the other party's incoming call as the malicious phone corresponding to the label selected by the user.
  • the method further includes:
  • the tag information is added to the voice of the recorded malicious phone according to the tag selected by the user.
  • the method further includes:
  • the method further includes:
  • the voice of the other party is sent to the cloud server, so that the cloud server performs voice analysis on the voice of the other party to obtain the voice feature point of the voice of the other party, and matches the voice feature point of the voice of the other party with the voice feature point in the voice feature database of the malicious phone, and matches A prompt to return a malicious call when successful;
  • the prompt of the malicious phone fed back by the cloud server is displayed on the call interface.
  • the embodiment of the invention further provides a method for configuring a voice feature database of a malicious phone, which is executed on the cloud server side, and includes:
  • the cloud server receives the voice of the malicious phone sent by the user equipment, and the voice of the malicious phone is a user device in the process of answering the phone, according to the user instruction, when detecting the user selecting the malicious phone option, the other party's call is identified as a malicious call, and Recording the voice of a malicious phone for a certain period of time;
  • the voice feature points of the malicious phone are saved in a malicious phone voice feature database.
  • the label of the malicious phone includes a label of the intermediary phone, a label of the promotion phone, a label of the scam phone, a label of the harassing phone, or a custom Telephone label
  • the voice analysis of the voice of the malicious phone is performed, and after obtaining the voice feature points of the malicious phone, the following includes:
  • the voice of the malicious phone carries the caller number of the malicious phone
  • the voice analysis of the voice of the malicious phone is performed, and after obtaining the voice feature points of the malicious phone, the following includes:
  • performing voice analysis on the voice of the malicious phone to obtain voice feature points of the malicious phone including:
  • the medium-average power spectrum of the noisy speech is obtained by means of frame averaging
  • a medium-term power spectrum based on the pure speech, a medium-duration power spectrum of the noisy speech, and A short-time power spectrum of noisy speech acquires a short-time power spectrum of pure speech
  • Equal loudness is aggravated by the short-term power spectrum of pure speech that eliminates multiplicative noise
  • the cloud server further includes:
  • the malicious phone voice feature database includes: caller number information of multiple malicious phones, voice feature point information of multiple malicious phones, tag information of voices of multiple malicious phones, and voice correspondence of multiple malicious phones. Textual information;
  • the malicious phone voice feature database further includes: a correspondence between a caller number of the malicious phone and a voice feature point of the malicious phone, a correspondence relationship between the voice feature point of the malicious phone and the tag, a voice feature point of the malicious phone, and text information.
  • a correspondence relationship the correspondence between the caller number of the malicious phone and the text message, the correspondence between the tag of the malicious phone and the text message, the one-to-one correspondence between the caller number and the voice feature point of the malicious phone, and the tag, the call of the malicious call A one-to-one correspondence between numbers, voice feature points, tags, and text information.
  • the method further includes:
  • the cloud server receives the voice of the other party collected by the user equipment during the process of answering the call;
  • the embodiment of the invention further provides a device for configuring a voice feature database of a malicious phone, which is located on the user equipment side, and includes:
  • the detecting module is configured to detect, according to a user instruction, a user selecting a malicious phone option during the answering call, and identifying the other party's incoming call as a malicious phone;
  • An acquisition module configured to record a voice of a malicious phone for a certain period of time
  • a sending module configured to send the voice of the malicious phone to the cloud server, so that the cloud server performs voice analysis on the voice of the malicious phone, obtains a voice feature point of the malicious phone, and saves the voice feature point of the malicious phone to the malicious
  • the phone voice feature database configured to store the voice of the malicious phone.
  • the malicious phone comprises an intermediary phone, a sales phone, a fraud phone, a harassing phone or a custom phone;
  • the detecting module is specifically configured to:
  • a phone label option is displayed on the call interface, including the label of the intermediary phone, the tag of the promotional phone, the tag of the scam phone, the tag of the harassing phone, or the option of the custom phone tag.
  • the user selects any label according to the voice of the other party in the phone label option, the user identifies the other party's incoming call as the malicious phone corresponding to the label selected by the user.
  • the device further includes:
  • Adding a module for adding the tag information to the voice of the recorded malicious phone according to the tag selected by the user Adding a module for adding the tag information to the voice of the recorded malicious phone according to the tag selected by the user.
  • the adding module is further configured to add a caller number in the voice of the recorded malicious phone.
  • the collecting module is further configured to collect a voice of the other party for a certain period of time during the process of answering the call;
  • the sending module is further configured to send the voice of the other party to the cloud server, so that the cloud server performs voice analysis on the voice of the other party, obtains a voice feature point of the voice of the other party, and the voice feature point of the voice of the other party and the voice feature database of the malicious phone The voice feature points are matched, and when the matching is successful, the prompt of the malicious phone is returned;
  • the display module is configured to display the prompt of the malicious phone on the call interface according to the prompt of the malicious phone fed back by the cloud server.
  • the embodiment of the invention provides a device for configuring a voice feature database of a malicious phone, which is located on the cloud server side, and includes:
  • a receiving module configured to receive a voice of a malicious phone sent by the user equipment, where the voice of the malicious phone is a user device in the process of answering the phone, according to the user instruction, when detecting the user selecting the malicious phone option, identifying the other party's incoming call as malicious Telephone, and record the voice of the malicious phone for a certain period of time;
  • a voice analysis module configured to perform voice analysis on voice of a malicious phone, and obtain a voice feature point of the malicious phone
  • a saving module configured to save the voice feature point of the malicious phone into a malicious phone voice feature database.
  • the label of the malicious phone includes a label of the intermediary phone, a label of the promotion phone, a label of the scam phone, a label of the harassing phone, or a custom Telephone label
  • the device further comprises:
  • Establishing a module configured to establish, according to the label carried in the voice of the malicious phone, a correspondence between a voice feature point of the malicious phone and the tag;
  • the saving module is further configured to save the correspondence between the voice feature point of the malicious phone and the tag to a malicious phone voice feature database.
  • the voice of the malicious phone carries the caller number of the malicious phone
  • the establishing module is further configured to establish, according to an incoming call number carried in the voice of the malicious phone, a correspondence between a voice feature point of the malicious phone and the caller number;
  • the saving module is further configured to save the correspondence between the voice feature point of the malicious phone and the caller number to a malicious phone voice feature database.
  • the voice analysis module is specifically configured to:
  • the frame-average method is used to obtain the noisy speech.
  • Equal loudness is aggravated by the short-term power spectrum of pure speech that eliminates multiplicative noise
  • the voice analysis module is further configured to identify the voice of the malicious phone as corresponding text information
  • the establishing module is further configured to establish a correspondence between the text information of the malicious phone and the voice feature point of the malicious phone.
  • the malicious phone voice feature database includes: caller number information of multiple malicious phones, voice feature point information of multiple malicious phones, tag information of voices of multiple malicious phones, and voice correspondence of multiple malicious phones. Textual information;
  • the malicious phone voice feature database further includes: a correspondence between a caller number of the malicious phone and a voice feature point of the malicious phone, a correspondence relationship between the voice feature point of the malicious phone and the tag, a voice feature point of the malicious phone, and text information.
  • a correspondence relationship the correspondence between the caller number of the malicious phone and the text message, the correspondence between the tag of the malicious phone and the text message, the one-to-one correspondence between the caller number and the voice feature point of the malicious phone, and the tag, the call of the malicious call A one-to-one correspondence between numbers, voice feature points, tags, and text information.
  • the receiving module is further configured to receive a voice of the partner that is collected by the user equipment during the time of answering the call;
  • the voice analysis module is further configured to perform voice analysis on the other party's voice to obtain the voice of the other party. Speech feature point
  • a matching module configured to match a voice feature point of the voice of the other party with a voice feature point in the voice feature database of the malicious phone
  • the sending module is configured to return a prompt of the malicious phone to the user equipment when the matching is successful, so that the user equipment displays the prompt of the malicious phone on the call interface according to the prompt of the malicious phone.
  • Embodiments of the present invention provide a computer program comprising computer readable code that, when run on a computing device, causes a computing device to perform the method of configuring a voice feature library of the malicious phone described above.
  • Embodiments of the present invention provide a computer readable medium in which the above computer program is stored.
  • the other party's incoming call is identified as a malicious call, and the voice of the malicious phone is recorded for a certain period of time;
  • Sending to the cloud server so that the cloud server performs voice analysis on the voice of the malicious phone, obtains a voice feature point of the malicious phone, and saves the voice feature point of the malicious phone to the malicious phone voice feature database.
  • the interception of a malicious phone can be achieved by matching the voice feature points.
  • FIG. 1 is a schematic flowchart of a method for configuring a voice feature database of a malicious phone according to an embodiment of the present application
  • FIG. 2 is a schematic structural diagram of a device for configuring a voice feature database of a malicious phone according to an embodiment of the present application
  • FIG. 3 is a schematic structural diagram of a device for configuring a voice feature library of a malicious phone according to an embodiment of the present application
  • FIG. 4 is a block diagram of a computing device for performing a configuration method of a voice feature library of a malicious phone according to an embodiment of the present invention
  • FIG. 5 is a storage unit of program code for holding or carrying a configuration method of a voice feature library implementing a malicious phone according to an embodiment of the present invention.
  • a computing device includes one or more processors (CPUs), input/output interfaces, network interfaces, and memory.
  • processors CPUs
  • input/output interfaces network interfaces
  • memory volatile and non-volatile memory
  • the memory may include non-persistent memory, random access memory (RAM), and/or non-volatile memory in a computer readable medium, such as read only memory (ROM) or flash memory.
  • RAM random access memory
  • ROM read only memory
  • Memory is an example of a computer readable medium.
  • Computer readable media includes both permanent and non-persistent, removable and non-removable media.
  • Information storage can be implemented by any method or technology.
  • the information can be computer readable instructions, data structures, modules of programs, or other data.
  • Examples of computer storage media include, but are not limited to, phase change memory (PRAM), static random access memory (SRAM), dynamic random access memory (DRAM), other types of random access memory (RAM), read only memory. (ROM), electrically erasable programmable read only memory (EEPROM), flash memory or other memory technology, compact disk read only memory (CD-ROM), digital versatile disk (DVD) or other optical storage, Magnetic tape cartridges, magnetic tape storage or other magnetic storage devices or any other non-transportable media can be used to store information that can be accessed by a computing device.
  • computer readable media does not include non-transitory computer readable media, such as modulated data signals and carrier waves.
  • a device can be directly electrically coupled to the second device, or electrically coupled to the second device indirectly through other devices or coupling means.
  • the method for configuring the voice feature database of the malicious phone is that, when the user answers the call, according to the user instruction, when the user selects the malicious phone option, the user is identified as a malicious call, and the voice of the malicious phone is performed. Recording for a certain period of time; sending the voice of the malicious phone to the cloud server, so that the cloud server performs voice analysis on the voice of the malicious phone, obtains the voice feature point of the malicious phone, and saves the voice feature point of the malicious phone to the malicious The phone voice feature database.
  • FIG. 1 is a schematic flowchart of a method for configuring a voice feature database of a malicious phone according to an embodiment of the present application. As shown in FIG. 1, the method includes:
  • the phone label option is displayed on the call interface
  • the malicious phone in this embodiment includes, for example, an intermediary phone, a sales phone, a fraud phone, a harassing phone, or a custom phone.
  • a corresponding phone tag is set for each malicious phone, wherein the phone tag option Examples include labels for intermediaries, labels for sales calls, labels for fraudulent calls, tags for harassing calls, or options for custom phone labels.
  • the electrical identifier is a malicious phone corresponding to the label selected by the user
  • the intermediary telephone label may be selected in the telephone label option to identify the other party's incoming call identification as an intermediary telephone; and, for example, when the user determines that the incoming call is a fraudulent telephone according to the voice of the opposite party, You can select the scam phone label in the phone label option to identify the other party's caller ID as a scam phone.
  • the method further includes: adding a phone label selected by the user to the voice of the recorded malicious phone according to the phone label selected by the user;
  • the method further includes: adding a caller number in the voice of the recorded malicious phone.
  • the cloud server performs voice analysis on the voice of the malicious phone, and obtains a voice feature point of the malicious phone;
  • the voice recognition feature may be extracted by using a voice recognition technology, which specifically includes:
  • the medium-average power spectrum of the noisy speech is obtained by means of frame averaging
  • the speech recognition technology of the present invention may be any existing speech recognition technology that can implement speech feature point lifting, and the present invention does not limit this.
  • the step 106 is specifically implemented to:
  • the step 106 is specifically implemented to:
  • the cloud server may further identify the voice of the malicious phone as corresponding text information; establish text information of the malicious phone and the Corresponding relationship between the voice feature points of the malicious phone, and storing the correspondence between the text information of the malicious phone and the voice feature point of the malicious phone in the malicious phone voice feature database.
  • the malicious phone voice feature database obtained by the method for configuring the voice feature database of the malicious phone according to the embodiment of the present invention includes, but is not limited to, caller number information of multiple malicious phones, and voice feature points of multiple malicious phones.
  • the malicious phone voice feature database may further include: a correspondence between a caller number of the malicious phone and a voice feature point of the malicious phone, a correspondence relationship between the voice feature point of the malicious phone and the tag, a voice feature point of the malicious phone, and text information.
  • a correspondence relationship the correspondence between the caller number of the malicious phone and the text message, the correspondence between the tag of the malicious phone and the text message, the one-to-one correspondence between the caller number and the voice feature point of the malicious phone, and the tag, the call of the malicious call A one-to-one correspondence between numbers, voice feature points, tags, and text information.
  • the malicious phone voice feature database obtained by the method for configuring the voice feature database of the malicious phone provided by the embodiment of the present invention can collect the voice of the other party for a certain period of time while the user answers the call; and send the voice of the other party to the cloud server,
  • the cloud server performs voice analysis on the voice of the other party to obtain a voice feature point of the voice of the other party, and matches the voice feature point of the voice of the other party with the voice feature point in the voice feature database of the malicious phone, and returns a prompt of the malicious phone when the matching is successful;
  • the prompt of the malicious phone fed back by the cloud server the prompt of the malicious phone is displayed on the call interface, so that the user hangs up the malicious call in time to prevent being deceived, thereby improving the security of the user using the phone.
  • FIG. 2 is a schematic structural diagram of a device for configuring a voice feature database of a malicious phone according to an embodiment of the present application, which is located on the user equipment side, as shown in FIG. 2, and includes:
  • the detecting module 21 is configured to detect, according to a user instruction, a user selecting a malicious phone option during the answering call, and identifying the other party's incoming call as a malicious phone;
  • the collecting module 22 is configured to record the voice of the malicious phone for a certain period of time
  • the sending module 23 is configured to send the voice of the malicious phone to the cloud server, so that the cloud server performs voice analysis on the voice of the malicious phone, obtains a voice feature point of the malicious phone, and saves the voice feature point of the malicious phone to Malicious phone voice feature database.
  • the malicious phone includes an intermediary phone, a sales phone, a fraud phone, a harassing phone or a custom phone;
  • the detecting module 21 is specifically configured to:
  • a phone label option is displayed on the call interface, including the label of the intermediary phone, the tag of the promotional phone, the tag of the scam phone, the tag of the harassing phone, or the option of the custom phone tag.
  • the user selects the phone label option according to the voice of the other party.
  • the other party's call is identified as the malicious call corresponding to the tag selected by the user.
  • the device further includes:
  • the adding module 24 is configured to add the label information to the voice of the recorded malicious phone according to the label selected by the user.
  • the adding module is further configured to add a caller number in the voice of the recorded malicious phone.
  • the collecting module 22 is further configured to collect a voice of the opposite party for a certain period of time during the process of answering the call;
  • the sending module 23 is further configured to send the voice of the other party to the cloud server, so that the cloud server performs voice analysis on the voice of the other party, obtains a voice feature point of the voice of the other party, and uses the voice feature point of the voice of the other party and the voice feature database of the malicious phone.
  • the voice feature points are matched, and when the matching is successful, the prompt of the malicious call is returned;
  • the device may further include:
  • the display module 25 is configured to display a prompt of the malicious phone on the call interface according to the prompt of the malicious phone fed back by the cloud server.
  • the apparatus shown in FIG. 2 can perform the method described in the embodiment shown in FIG. 1, and the implementation principle and technical effects are not described again.
  • FIG. 3 is a schematic structural diagram of a device for configuring a voice feature database of a malicious phone according to an embodiment of the present application, which is located on the cloud server side, as shown in FIG. 3, and includes:
  • the receiving module 31 is configured to receive the voice of the malicious phone sent by the user equipment, where the voice of the malicious phone is a user equipment in the process of answering the phone, according to the user instruction, when detecting the user selecting the malicious phone option, identifying the other party's incoming call as Malicious calls, and the voice of the malicious phone is recorded for a certain period of time;
  • the voice analysis module 32 is configured to perform voice analysis on the voice of the malicious phone to obtain a voice feature point of the malicious phone;
  • the saving module 33 is configured to save the voice feature points of the malicious phone into the malicious phone voice feature database.
  • the malicious phone If the voice of the malicious phone carries the label of the malicious phone; the malicious phone The label includes the label of the intermediary phone, the tag of the promotional phone, the tag of the scam phone, the tag of the harassing phone, or a custom phone tag;
  • the device further comprises:
  • the establishing module 34 is configured to establish, according to the label carried in the voice of the malicious phone, a correspondence between a voice feature point of the malicious phone and the tag;
  • the saving module 33 is further configured to save the correspondence between the voice feature point of the malicious phone and the tag to a malicious phone voice feature database.
  • the establishing module 34 is further configured to establish, according to the caller number carried in the voice of the malicious phone, a correspondence between a voice feature point of the malicious phone and the caller number;
  • the saving module 33 is further configured to save the correspondence between the voice feature point of the malicious phone and the caller number to the malicious phone voice feature database.
  • the voice analysis module 32 is specifically configured to:
  • the medium-average power spectrum of the noisy speech is obtained by means of frame averaging
  • Equal loudness is aggravated by the short-term power spectrum of pure speech that eliminates multiplicative noise
  • the voice analysis module 32 is further configured to identify the voice of the malicious phone as corresponding text information
  • the establishing module 34 is further configured to establish a correspondence between the text information of the malicious phone and the voice feature point of the malicious phone.
  • the malicious phone voice feature database includes: caller number information of multiple malicious phones, voice feature point information of multiple malicious phones, tag information of voices of multiple malicious phones, and multiple malicious phones. Text information corresponding to the voice;
  • the malicious phone voice feature database further includes: a correspondence between a caller number of the malicious phone and a voice feature point of the malicious phone, a correspondence relationship between the voice feature point of the malicious phone and the tag, a voice feature point of the malicious phone, and text information.
  • a correspondence relationship the correspondence between the caller number of the malicious phone and the text message, the correspondence between the tag of the malicious phone and the text message, the one-to-one correspondence between the caller number and the voice feature point of the malicious phone, and the tag, the call of the malicious call A one-to-one correspondence between numbers, voice feature points, tags, and text information.
  • the receiving module 31 is further configured to receive a voice of the partner that is collected by the user equipment during the answering process;
  • the voice analysis module 32 is further configured to perform voice analysis on the voice of the other party to obtain a voice feature point of the voice of the other party;
  • the matching module 35 is configured to match the voice feature points of the counterpart voice with the voice feature points in the malicious phone voice feature database
  • the sending module 36 is configured to return a prompt of the malicious phone to the user equipment when the matching is successful, so that the user equipment displays the prompt of the malicious phone on the call interface according to the prompt of the malicious phone.
  • the apparatus shown in FIG. 3 can perform the method described in the embodiment shown in FIG. 1, and the implementation principle and technical effects are not described again.
  • the embodiment of the present invention further provides a system for configuring a voice feature database of a malicious phone, including a user equipment and a cloud server;
  • the user equipment includes the configuration device shown in FIG. 2, and the cloud server includes the configuration device shown in FIG. 3, and details are not described herein.
  • Storage 4 shows a computing device that can implement a method of configuring a voice feature library of a malicious phone in accordance with the present invention.
  • the computing device conventionally includes a processor 410 and a computer program product or computer readable medium in the form of a storage device 420.
  • Storage device 420 can be an electronic memory such as a flash memory, EEPROM (Electrically Erasable Programmable Read Only Memory), EPROM, hard disk, or ROM.
  • Storage device 420 has a storage space 430 that stores program code 431 for performing any of the method steps described above.
  • storage space 430 storing program code may include various program code 431 for implementing various steps in the above methods, respectively.
  • the program code can be read from or written to one or more computer program products.
  • These computer program products include program code carriers such as a hard disk, a compact disk (CD), a memory card, or a floppy disk.
  • Such computer program products are typically portable or fixed storage units such as those shown in FIG.
  • the storage unit may have storage segments, storage spaces, and the like that are similarly arranged to storage device 420 in the computing device of FIG.
  • the program code can be compressed, for example, in an appropriate form.
  • the storage unit comprises computer readable code 431' for performing the steps of the method according to the invention, ie code that can be read by a processor such as 410, which when executed by the computing device causes the computing device Perform the various steps in the method described above.

Landscapes

  • Engineering & Computer Science (AREA)
  • Signal Processing (AREA)
  • Computer Security & Cryptography (AREA)
  • Acoustics & Sound (AREA)
  • Health & Medical Sciences (AREA)
  • Audiology, Speech & Language Pathology (AREA)
  • Human Computer Interaction (AREA)
  • Physics & Mathematics (AREA)
  • Multimedia (AREA)
  • Computational Linguistics (AREA)
  • Computer Vision & Pattern Recognition (AREA)
  • Telephonic Communication Services (AREA)
  • Telephone Function (AREA)

Abstract

本发明公开了一种恶意电话的语音特征库的配置方法和装置,在接听电话过程中,根据用户指令,在检测到用户选择恶意电话选项时,将对方来电标识为恶意电话,并对恶意电话的语音进行一定时长的录音;将恶意电话的语音发送给云端服务器,以使所述云端服务器对恶意电话的语音进行语音分析,得到恶意电话的语音特征点,将所述恶意电话的语音特征点保存到恶意电话语音特征数据库中。

Description

恶意电话的语音特征库的配置方法和装置
相关申请的交叉参考
本申请主张在2015年11月20日在中国提交的中国专利申请号No.201510808778.x的优先权,其全部内容通过引用包含于此。
技术领域
本发明属于移动设备技术领域,具体地说,涉及一种恶意电话的语音特征库的配置方法和装置。
背景技术
当前社会上利用电话不经受话方许可,进行强制性非法电话广告、电话宣传和传真,他们不但在白天也经常在深夜,连续不断地拨打骚扰电话,肆无忌惮地干扰人们正常的生活环境和工作环境,由于骚扰电话很多是通过网络电话的方法实施骚扰话动,现有的公共电话网又无法识别骚扰电话的来源,也很难分辨骚扰电话和正常电话,而且由于当前的电话机只能通过由用户自己预置信任电话号码,隔离其它所有电话的方法来防止骚扰电话,这就大大限制了电话机的用途,因此骚扰电话很难杜绝。电话已被社会上破坏分子及不良的商业竞争所利用,进行不法骚扰活动。
发明内容
有鉴于此,本申请提供了一种恶意电话的语音特征库的配置方法和装置,可以实现通过语音特征点的匹配对恶意电话的拦截。
本发明实施例提供一种恶意电话的语音特征库的配置方法,位于用户设备侧执行,包括:
在接听电话过程中,根据用户指令,在检测到用户选择恶意电话选项时,将对方来电标识为恶意电话,并对恶意电话的语音进行一定时长的录音;
将恶意电话的语音发送给云端服务器,以使所述云端服务器对恶意电话 的语音进行语音分析,得到恶意电话的语音特征点,将所述恶意电话的语音特征点保存到恶意电话语音特征数据库中。
可选地,所述恶意电话包括中介电话、推销电话、诈骗电话、骚扰电话或自定义电话;
在接听电话过程中,根据用户指令,在检测到用户选择恶意电话选项时,将对方来电标识为恶意电话,包括:
在接听电话过程中,在通话界面上显示电话标签选项,所述电话标签选项中包括中介电话的标签、推销电话的标签、诈骗电话的标签、骚扰电话的标签或自定义电话标签的选项,以便用户根据对方语音在电话标签选项中选择任一标签时,将对方来电标识为用户选择的标签所对应的恶意电话。
可选地,对恶意电话的语音进行一定时长的录音之后,还包括:
根据用户选择的标签,在录制的恶意电话的语音中添加所述标签信息。
可选地,对恶意电话的语音进行一定时长的录音之后,还包括:
在录制的恶意电话的语音中添加来电号码。
可选地,所述的方法还包括:
在接听电话过程中,采集一定时长的对方语音;
将对方语音发送给云端服务器,以使云端服务器将对方语音进行语音分析,得到对方语音的语音特征点,将对方语音的语音特征点与恶意电话语音特征数据库中的语音特征点进行匹配,在匹配成功时,返回恶意电话的提示;
根据云端服务器反馈的恶意电话的提示,在通话界面上显示所述恶意电话的提示。
本发明实施例还提供一种恶意电话的语音特征库的配置方法,位于云端服务器侧执行,包括:
云端服务器接收用户设备发送的恶意电话的语音,所述恶意电话的语音为用户设备在接听电话过程中,根据用户指令,在检测到用户选择恶意电话选项时,将对方来电标识为恶意电话,并对恶意电话的语音进行一定时长的录音;
对恶意电话的语音进行语音分析,得到恶意电话的语音特征点;
将所述恶意电话的语音特征点保存到恶意电话语音特征数据库中。
可选地,若所述恶意电话的语音中携带有所述恶意电话的标签;所述恶意电话的标签包括中介电话的标签、推销电话的标签、诈骗电话的标签、骚扰电话的标签或自定义电话标签;
则对恶意电话的语音进行语音分析,得到恶意电话的语音特征点之后,包括:
根据所述恶意电话的语音中携带的标签,建立所述恶意电话的语音特征点与所述标签的对应关系;
将所述恶意电话的语音特征点与所述标签的对应关系保存到恶意电话语音特征数据库。
可选地,若所述恶意电话的语音中携带有所述恶意电话的来电号码;
则对恶意电话的语音进行语音分析,得到恶意电话的语音特征点之后,包括:
根据所述恶意电话的语音中携带的来电号码,建立所述恶意电话的语音特征点与所述来电号码的对应关系;
将所述恶意电话的语音特征点与所述来电号码的对应关系保存到恶意电话语音特征数据库。
可选地,对恶意电话的语音进行语音分析,得到恶意电话的语音特征点,包括:
获取所述恶意电话的语音的功率谱;
将所获得的功率谱通过滤波处理,获得含噪语音的短时功率谱;
根据所获得的含噪语音的短时功率谱,采用帧平均的方式求取含噪语音的中等时长功率谱;
对所获得的含噪语音的中等时长功率谱进行不对称滤波和掩蔽抗噪,以得到纯净语音的中等时长功率谱;
根据所述纯净语音的中等时长功率谱、含噪语音的中等时长功率谱和含 噪语音的短时功率谱获取纯净语音的短时功率谱;
对纯净语音的短时功率谱进行能量归一化处理,以消除乘性噪音;
对消除了乘性噪音的纯净语音的短时功率谱进行等响度加重
对等响度加重后的纯净语音的短时功率谱进行指数非线性操作;
对进行了指数非线性操作后的纯净语音的短时功率谱进行傅立叶逆变换,以求取倒谱系数;
对倒谱系数进行均值归一化处理,得到所述恶意电话的语音特征点。
可选地,云端服务器接收用户设备发送的恶意电话的语音之后,还包括:
将所述恶意电话的语音识别成对应的文字信息;
建立所述恶意电话的文字信息与所述恶意电话的语音特征点之间的对应关系。
可选地,所述恶意电话语音特征数据库中包括:多个恶意电话的来电号码信息、多个恶意电话的语音特征点信息、多个恶意电话的语音的标签信息、多个恶意电话的语音对应的文字信息;
所述恶意电话语音特征数据库中还包括:恶意电话的来电号码与恶意电话的语音特征点之间的对应关系、恶意电话的语音特征点与标签的对应关系、恶意电话的语音特征点与文字信息的对应关系、恶意电话的来电号码与文字信息的对应关系、恶意电话的标签与文字信息的对应关系、恶意电话的来电号码和语音特征点以及标签之间的一一对应关系、恶意电话的来电号码、语音特征点、标签以及文字信息之间的一一对应关系。
可选地,所述的方法还包括:
云端服务器接收用户设备在接听电话过程中采集的一定时长的对方语音;
将对方语音进行语音分析,得到对方语音的语音特征点;
将对方语音的语音特征点与恶意电话语音特征数据库中的语音特征点进行匹配;
在匹配成功时,向用户设备返回恶意电话的提示以使用户设备根据所述 恶意电话的提示,在通话界面上显示所述恶意电话的提示。
本发明实施例还提供一种恶意电话的语音特征库的配置装置,位于用户设备侧,包括:
检测模块,用于在接听电话过程中,根据用户指令,检测到用户选择恶意电话选项,将对方来电标识为恶意电话,;
采集模块,用于对恶意电话的语音进行一定时长的录音;
发送模块,用于将恶意电话的语音发送给云端服务器,以使所述云端服务器对恶意电话的语音进行语音分析,得到恶意电话的语音特征点,将所述恶意电话的语音特征点保存到恶意电话语音特征数据库中。
可选地,所述恶意电话包括中介电话、推销电话、诈骗电话、骚扰电话或自定义电话;
所述检测模块具体用于:
在接听电话过程中,在通话界面上显示电话标签选项,所述电话标签选项中包括中介电话的标签、推销电话的标签、诈骗电话的标签、骚扰电话的标签或自定义电话标签的选项,以便用户根据对方语音在电话标签选项中选择任一标签时,将对方来电标识为用户选择的标签所对应的恶意电话。
可选地,所述的装置还包括:
添加模块,用于根据用户选择的标签,在录制的恶意电话的语音中添加所述标签信息。
所述添加模块,还用于在录制的恶意电话的语音中添加来电号码。
所述采集模块,还用于在接听电话过程中,采集一定时长的对方语音;
所述发送模块,还用于将对方语音发送给云端服务器,以使云端服务器将对方语音进行语音分析,得到对方语音的语音特征点,将对方语音的语音特征点与恶意电话语音特征数据库中的语音特征点进行匹配,在匹配成功时,返回恶意电话的提示;
显示模块,用于根据云端服务器反馈的恶意电话的提示,在通话界面上显示所述恶意电话的提示。
本发明实施例提供一种恶意电话的语音特征库的配置装置,位于云端服务器侧,包括:
接收模块,用于接收用户设备发送的恶意电话的语音,所述恶意电话的语音为用户设备在接听电话过程中,根据用户指令,在检测到用户选择恶意电话选项时,将对方来电标识为恶意电话,并对恶意电话的语音进行一定时长的录音;
语音分析模块,用于对恶意电话的语音进行语音分析,得到恶意电话的语音特征点;
保存模块,用于将所述恶意电话的语音特征点保存到恶意电话语音特征数据库中。
可选地,若所述恶意电话的语音中携带有所述恶意电话的标签;所述恶意电话的标签包括中介电话的标签、推销电话的标签、诈骗电话的标签、骚扰电话的标签或自定义电话标签;
则所述装置还包括:
建立模块,用于根据所述恶意电话的语音中携带的标签,建立所述恶意电话的语音特征点与所述标签的对应关系;
所述保存模块,还用于将所述恶意电话的语音特征点与所述标签的对应关系保存到恶意电话语音特征数据库。
可选地,若所述恶意电话的语音中携带有所述恶意电话的来电号码;
所述建立模块,还用于根据所述恶意电话的语音中携带的来电号码,建立所述恶意电话的语音特征点与所述来电号码的对应关系;
所述保存模块,还用于将所述恶意电话的语音特征点与所述来电号码的对应关系保存到恶意电话语音特征数据库。
可选地,所述语音分析模块具体用于:
获取所述恶意电话的语音的功率谱;
将所获得的功率谱通过滤波处理,获得含噪语音的短时功率谱;
根据所获得的含噪语音的短时功率谱,采用帧平均的方式求取含噪语音 的中等时长功率谱;
对所获得的含噪语音的中等时长功率谱进行不对称滤波和掩蔽抗噪,以得到纯净语音的中等时长功率谱;
根据所述纯净语音的中等时长功率谱、含噪语音的中等时长功率谱和含噪语音的短时功率谱获取纯净语音的短时功率谱;
对纯净语音的短时功率谱进行能量归一化处理,以消除乘性噪音;
对消除了乘性噪音的纯净语音的短时功率谱进行等响度加重
对等响度加重后的纯净语音的短时功率谱进行指数非线性操作;
对进行了指数非线性操作后的纯净语音的短时功率谱进行傅立叶逆变换,以求取倒谱系数;
对倒谱系数进行均值归一化处理,得到所述恶意电话的语音特征点。
可选地,所述语音分析模块,还用于将所述恶意电话的语音识别成对应的文字信息;
所述建立模块,还用于建立所述恶意电话的文字信息与所述恶意电话的语音特征点之间的对应关系。
可选地,所述恶意电话语音特征数据库中包括:多个恶意电话的来电号码信息、多个恶意电话的语音特征点信息、多个恶意电话的语音的标签信息、多个恶意电话的语音对应的文字信息;
所述恶意电话语音特征数据库中还包括:恶意电话的来电号码与恶意电话的语音特征点之间的对应关系、恶意电话的语音特征点与标签的对应关系、恶意电话的语音特征点与文字信息的对应关系、恶意电话的来电号码与文字信息的对应关系、恶意电话的标签与文字信息的对应关系、恶意电话的来电号码和语音特征点以及标签之间的一一对应关系、恶意电话的来电号码、语音特征点、标签以及文字信息之间的一一对应关系。
所述接收模块,还用于接收用户设备在接听电话过程中采集的一定时长的对方语音;
所述语音分析模块,还用于将对方语音进行语音分析,得到对方语音的 语音特征点;
匹配模块,用于将对方语音的语音特征点与恶意电话语音特征数据库中的语音特征点进行匹配;
所述发送模块,用于在匹配成功时,向用户设备返回恶意电话的提示以使用户设备根据所述恶意电话的提示,在通话界面上显示所述恶意电话的提示。
本发明实施例提供一种计算机程序,其包括计算机可读代码,当计算机可读代码在计算设备上运行时,导致计算设备执行上述的恶意电话的语音特征库的配置方法。
本发明实施例提供了一种计算机可读介质,其中存储了上述计算机程序。
本发明实施例通过在接听电话过程中,根据用户指令,在检测到用户选择恶意电话选项时,将对方来电标识为恶意电话,并对恶意电话的语音进行一定时长的录音;将恶意电话的语音发送给云端服务器,以使所述云端服务器对恶意电话的语音进行语音分析,得到恶意电话的语音特征点,将所述恶意电话的语音特征点保存到恶意电话语音特征数据库中。可以通过语音特征点的匹配实现对恶意电话的拦截。
附图概述
此处所说明的附图用来提供对本申请的进一步理解,构成本申请的一部分,本申请的示意性实施例及其说明用于解释本申请,并不构成对本申请的不当限定。在附图中:
图1是本申请实施例的提供的一种恶意电话的语音特征库的配置方法的流程示意图;
图2是本申请实施例的提供的一种恶意电话的语音特征库的配置装置的结构示意图;
图3是本申请实施例的提供的一种恶意电话的语音特征库的配置装置的结构示意图;
图4是本申请实施例的用于执行根据本发明实施例的恶意电话的语音特征库的配置方法的计算设备的框图;
图5是本申请实施例的用于保持或者携带实现根据本发明实施例的恶意电话的语音特征库的配置方法的程序代码的存储单元。
本发明的较佳实施方式
以下将配合附图及实施例来详细说明本发明的实施方式,藉此对本发明如何应用技术手段来解决技术问题并达成技术功效的实现过程能充分理解并据以实施。
在一个典型的配置中,计算设备包括一个或多个处理器(CPU)、输入/输出接口、网络接口和内存。
内存可能包括计算机可读介质中的非永久性存储器,随机存取存储器(RAM)和/或非易失性内存等形式,如只读存储器(ROM)或闪存(flash RAM)。内存是计算机可读介质的示例。
计算机可读介质包括永久性和非永久性、可移动和非可移动媒体可以由任何方法或技术来实现信息存储。信息可以是计算机可读指令、数据结构、程序的模块或其他数据。计算机的存储介质的例子包括,但不限于相变内存(PRAM)、静态随机存取存储器(SRAM)、动态随机存取存储器(DRAM)、其他类型的随机存取存储器(RAM)、只读存储器(ROM)、电可擦除可编程只读存储器(EEPROM)、快闪记忆体或其他内存技术、只读光盘只读存储器(CD-ROM)、数字多功能光盘(DVD)或其他光学存储、磁盒式磁带,磁带磁磁盘存储或其他磁性存储设备或任何其他非传输介质,可用于存储可以被计算设备访问的信息。按照本文中的界定,计算机可读介质不包括非暂存电脑可读媒体(transitory media),如调制的数据信号和载波。
如在说明书及权利要求当中调用了某些词汇来指称特定组件。本领域技术人员应可理解,硬件制造商可能会用不同名词来称呼同一个组件。本说明书及权利要求并不以名称的差异来作为区分组件的方式,而是以组件在功能上的差异来作为区分的准则。如在通篇说明书及权利要求当中所提及的“包 含”为一开放式用语,故应解释成“包含但不限定于”。“大致”是指在可接收的误差范围内,本领域技术人员能够在一定误差范围内解决所述技术问题,基本达到所述技术效果。此外,“耦接”一词在此包含任何直接及间接的电性耦接手段。因此,若文中描述一第一装置耦接于一第二装置,则代表所述第一装置可直接电性耦接于所述第二装置,或通过其他装置或耦接手段间接地电性耦接至所述第二装置。说明书后续描述为实施本发明的较佳实施方式,然所述描述乃以说明本发明的一般原则为目的,并非用以限定本发明的范围。本发明的保护范围当视所附权利要求所界定者为准。
还需要说明的是,术语“包括”、“包含”或者其任何其他变体意在涵盖非排他性的包含,从而使得包括一系列要素的商品或者系统不仅包括那些要素,而且还包括没有明确列出的其他要素,或者是还包括为这种商品或者系统所固有的要素。在没有更多限制的情况下,由语句“包括一个……”限定的要素,并不排除在包括所述要素的商品或者系统中还存在另外的相同要素
本发明提供的恶意电话的语音特征库的配置方法是用户在接听电话过程中,根据用户指令,在检测到用户选择恶意电话选项时,将对方来电标识为恶意电话,并对恶意电话的语音进行一定时长的录音;将恶意电话的语音发送给云端服务器,以使所述云端服务器对恶意电话的语音进行语音分析,得到恶意电话的语音特征点,将所述恶意电话的语音特征点保存到恶意电话语音特征数据库中。
以下通过具体实施例对本发明的技术方案进行详细的描述。
图1是本申请实施例的提供的一种恶意电话的语音特征库的配置方法的流程示意图,如图1所示,包括:
101、接听电话过程中,在通话界面上显示电话标签选项;
本实施例的恶意电话例如包括中介电话、推销电话、诈骗电话、骚扰电话或自定义电话等,相对应,本实施例中,对每一恶意电话设置有对应的电话标签,其中,电话标签选项例如包括中介电话的标签、推销电话的标签、诈骗电话的标签、骚扰电话的标签或自定义电话标签的选项。
102、用户根据对方语音在电话标签选项中选择任一标签时,将对方来 电标识为用户选择的标签所对应的恶意电话;
具体地,例如,当用户根据对方语音确定来电是中介电话,可以在电话标签选项中选择中介电话标签,将对方来电标识标识为中介电话;又例如,当用户根据对方语音确定来电是诈骗电话,可以在电话标签选项中选择诈骗电话标签,将对方来电标识标识为诈骗电话,这里不一一举例说明。
103、对恶意电话的语音进行一定时长的录音;
可选地,步骤103中对恶意电话的语音进行一定时长的录音之后,还包括:根据用户选择的电话标签,在录制的恶意电话的语音中添加用户选择的电话标签;
可选地,步骤103中对恶意电话的语音进行一定时长的录音之后,还包括:在录制的恶意电话的语音中添加来电号码。
104、将恶意电话的语音发送给云端服务器;
105、云端服务器对恶意电话的语音进行语音分析,得到恶意电话的语音特征点;
具体地,步骤105具体实现时可以利用语音识别技术进行语音特征点的提取,具体包括:
获取所述恶意电话的语音的功率谱;
将所获得的功率谱通过滤波处理,获得含噪语音的短时功率谱;
根据所获得的含噪语音的短时功率谱,采用帧平均的方式求取含噪语音的中等时长功率谱;
对所获得的含噪语音的中等时长功率谱进行不对称滤波和掩蔽抗噪,以得到纯净语音的中等时长功率谱;
根据所述纯净语音的中等时长功率谱、含噪语音的中等时长功率谱和含噪语音的短时功率谱获取纯净语音的短时功率谱;
对纯净语音的短时功率谱进行能量归一化处理,以消除乘性噪音;
对消除了乘性噪音的纯净语音的短时功率谱进行等响度加重;
对等响度加重后的纯净语音的短时功率谱进行指数非线性操作;
对进行了指数非线性操作后的纯净语音的短时功率谱进行傅立叶逆变换,以求取倒谱系数;
对倒谱系数进行均值归一化处理,得到所述恶意电话的语音特征点。
本发明所述的语音识别技术可以是现有的任何可以实现语音特征点提起的语音识别技术,本发明对此不作任何限定。
106、将恶意电话的语音特征点保存到恶意电话语音特征数据库中。
在一种可选的实施方式中,若所述恶意电话的语音中携带有所述恶意电话的标签;则步骤106在具体实现时包括:
根据所述恶意电话的语音中携带的标签,建立所述恶意电话的语音特征点与所述标签的对应关系;
将所述恶意电话的语音特征点与所述标签的对应关系保存到恶意电话语音特征数据库。
在一种可选的实施方式中,若所述恶意电话的语音中携带有所述恶意电话的来电号码;则步骤106在具体实现时包括:
根据所述恶意电话的语音中携带的来电号码,建立所述恶意电话的语音特征点与所述来电号码的对应关系;
将所述恶意电话的语音特征点与所述来电号码的对应关系保存到恶意电话语音特征数据库。
在一种可选的实施方式中,云端服务器接收用户设备发送的恶意电话的语音之后,还可以将所述恶意电话的语音识别成对应的文字信息;建立所述恶意电话的文字信息与所述恶意电话的语音特征点之间的对应关系,并将所述恶意电话的文字信息与所述恶意电话的语音特征点之间的对应关系保存到恶意电话语音特征数据库中。
由此可知,根据本发明实施例提供的恶意电话的语音特征库的配置方法得到的恶意电话语音特征数据库中包括但不限于:多个恶意电话的来电号码信息、多个恶意电话的语音特征点信息、多个恶意电话的语音的标签信息、多个恶意电话的语音对应的文字信息;
进一步地,本发明实施例提供的恶意电话的语音特征库的配置方法得到 的恶意电话语音特征数据库中还可以包括:恶意电话的来电号码与恶意电话的语音特征点之间的对应关系、恶意电话的语音特征点与标签的对应关系、恶意电话的语音特征点与文字信息的对应关系、恶意电话的来电号码与文字信息的对应关系、恶意电话的标签与文字信息的对应关系、恶意电话的来电号码和语音特征点以及标签之间的一一对应关系、恶意电话的来电号码、语音特征点、标签以及文字信息之间的一一对应关系。
基于本发明实施例提供的恶意电话的语音特征库的配置方法得到的恶意电话语音特征数据库,当用户在接听电话过程中,可以采集一定时长的对方语音;并将对方语音发送给云端服务器,以使云端服务器将对方语音进行语音分析,得到对方语音的语音特征点,将对方语音的语音特征点与恶意电话语音特征数据库中的语音特征点进行匹配,在匹配成功时,返回恶意电话的提示;进一步地,根据云端服务器反馈的恶意电话的提示,在通话界面上显示所述恶意电话的提示,以便用户及时挂断电话恶意电话,防止上当受骗,因此可以提高用户使用电话的安全性。
图2是本申请实施例的提供的一种恶意电话的语音特征库的配置装置的结构示意图,位于用户设备侧,如图2所示,包括:
检测模块21,用于在接听电话过程中,根据用户指令,检测到用户选择恶意电话选项,将对方来电标识为恶意电话,;
采集模块22,用于对恶意电话的语音进行一定时长的录音;
发送模块23,用于将恶意电话的语音发送给云端服务器,以使所述云端服务器对恶意电话的语音进行语音分析,得到恶意电话的语音特征点,将所述恶意电话的语音特征点保存到恶意电话语音特征数据库中。
其中,所述恶意电话包括中介电话、推销电话、诈骗电话、骚扰电话或自定义电话;
所述检测模块21具体用于:
在接听电话过程中,在通话界面上显示电话标签选项,所述电话标签选项中包括中介电话的标签、推销电话的标签、诈骗电话的标签、骚扰电话的标签或自定义电话标签的选项,以便用户根据对方语音在电话标签选项中选 择任一标签时,将对方来电标识为用户选择的标签所对应的恶意电话。
可选地,所述的装置还包括:
添加模块24,用于根据用户选择的标签,在录制的恶意电话的语音中添加所述标签信息。
所述添加模块,还用于在录制的恶意电话的语音中添加来电号码。
其中,可选地,所述采集模块22,还用于在接听电话过程中,采集一定时长的对方语音;
所述发送模块23,还用于将对方语音发送给云端服务器,以使云端服务器将对方语音进行语音分析,得到对方语音的语音特征点,将对方语音的语音特征点与恶意电话语音特征数据库中的语音特征点进行匹配,在匹配成功时,返回恶意电话的提示;
所述装置还可以包括:
显示模块25,用于根据云端服务器反馈的恶意电话的提示,在通话界面上显示所述恶意电话的提示。
图2所示装置可以执行图1所示实施例所述的方法,其实现原理和技术效果不再赘述。
图3是本申请实施例的提供的一种恶意电话的语音特征库的配置装置的结构示意图,位于云端服务器侧,如图3所示,包括:
接收模块31,用于接收用户设备发送的恶意电话的语音,所述恶意电话的语音为用户设备在接听电话过程中,根据用户指令,在检测到用户选择恶意电话选项时,将对方来电标识为恶意电话,并对恶意电话的语音进行一定时长的录音;
语音分析模块32,用于对恶意电话的语音进行语音分析,得到恶意电话的语音特征点;
保存模块33,用于将所述恶意电话的语音特征点保存到恶意电话语音特征数据库中。
若所述恶意电话的语音中携带有所述恶意电话的标签;所述恶意电话的 标签包括中介电话的标签、推销电话的标签、诈骗电话的标签、骚扰电话的标签或自定义电话标签;
则所述装置还包括:
建立模块34,用于根据所述恶意电话的语音中携带的标签,建立所述恶意电话的语音特征点与所述标签的对应关系;
所述保存模块33,还用于将所述恶意电话的语音特征点与所述标签的对应关系保存到恶意电话语音特征数据库。
若所述恶意电话的语音中携带有所述恶意电话的来电号码;
所述建立模块34,还用于根据所述恶意电话的语音中携带的来电号码,建立所述恶意电话的语音特征点与所述来电号码的对应关系;
所述保存模块33,还用于将所述恶意电话的语音特征点与所述来电号码的对应关系保存到恶意电话语音特征数据库。
可选地,所述语音分析模块32具体用于:
获取所述恶意电话的语音的功率谱;
将所获得的功率谱通过滤波处理,获得含噪语音的短时功率谱;
根据所获得的含噪语音的短时功率谱,采用帧平均的方式求取含噪语音的中等时长功率谱;
对所获得的含噪语音的中等时长功率谱进行不对称滤波和掩蔽抗噪,以得到纯净语音的中等时长功率谱;
根据所述纯净语音的中等时长功率谱、含噪语音的中等时长功率谱和含噪语音的短时功率谱获取纯净语音的短时功率谱;
对纯净语音的短时功率谱进行能量归一化处理,以消除乘性噪音;
对消除了乘性噪音的纯净语音的短时功率谱进行等响度加重
对等响度加重后的纯净语音的短时功率谱进行指数非线性操作;
对进行了指数非线性操作后的纯净语音的短时功率谱进行傅立叶逆变换,以求取倒谱系数;
对倒谱系数进行均值归一化处理,得到所述恶意电话的语音特征点。
所述语音分析模块32,还用于将所述恶意电话的语音识别成对应的文字信息;
所述建立模块34,还用于建立所述恶意电话的文字信息与所述恶意电话的语音特征点之间的对应关系。
本发明实施例中,所述恶意电话语音特征数据库中包括:多个恶意电话的来电号码信息、多个恶意电话的语音特征点信息、多个恶意电话的语音的标签信息、多个恶意电话的语音对应的文字信息;
所述恶意电话语音特征数据库中还包括:恶意电话的来电号码与恶意电话的语音特征点之间的对应关系、恶意电话的语音特征点与标签的对应关系、恶意电话的语音特征点与文字信息的对应关系、恶意电话的来电号码与文字信息的对应关系、恶意电话的标签与文字信息的对应关系、恶意电话的来电号码和语音特征点以及标签之间的一一对应关系、恶意电话的来电号码、语音特征点、标签以及文字信息之间的一一对应关系。
可选地,所述接收模块31,还用于接收用户设备在接听电话过程中采集的一定时长的对方语音;
所述语音分析模块32,还用于将对方语音进行语音分析,得到对方语音的语音特征点;
匹配模块35,用于将对方语音的语音特征点与恶意电话语音特征数据库中的语音特征点进行匹配;
所述发送模块36,用于在匹配成功时,向用户设备返回恶意电话的提示以使用户设备根据所述恶意电话的提示,在通话界面上显示所述恶意电话的提示。
图3所示装置可以执行图1所示实施例所述的方法,其实现原理和技术效果不再赘述。
本发明实施例还提供的一种恶意电话的语音特征库的配置系统,包括用户设备和云端服务器;
其中,用户设备中包括图2所示的配置装置,云端服务器包括图3所示的配置装置,具体不再赘述。
图4示出了可以实现根据本发明的恶意电话的语音特征库的配置方法的计算设备。该计算设备传统上包括处理器410和以存储设备420形式的计算机程序产品或者计算机可读介质。存储设备420可以是诸如闪存、EEPROM(电可擦除可编程只读存储器)、EPROM、硬盘或者ROM之类的电子存储器。存储设备420具有存储用于执行上述方法中的任何方法步骤的程序代码431的存储空间430。例如,存储程序代码的存储空间430可以包括分别用于实现上面的方法中的各种步骤的各个程序代码431。这些程序代码可以从一个或者多个计算机程序产品中读出或者写入到这一个或者多个计算机程序产品中。这些计算机程序产品包括诸如硬盘、紧致盘(CD)、存储卡或者软盘之类的程序代码载体。这样的计算机程序产品通常为例如图5所示的便携式或者固定存储单元。该存储单元可以具有与图4的计算设备中的存储设备420类似布置的存储段、存储空间等。程序代码可以例如以适当形式进行压缩。通常,存储单元包括用于执行根据本发明的方法步骤的计算机可读代码431',即可以由诸如410之类的处理器读取的代码,当这些代码由计算设备运行时,导致该计算设备执行上面所描述的方法中的各个步骤。
上述说明示出并描述了本发明的若干优选实施例,但如前所述,应当理解本发明并非局限于本文所披露的形式,不应看作是对其他实施例的排除,而可用于各种其他组合、修改和环境,并能够在本文所述发明构想范围内,通过上述教导或相关领域的技术或知识进行改动。而本领域人员所进行的改动和变化不脱离本发明的精神和范围,则都应在本发明所附权利要求的保护范围内。

Claims (26)

  1. 一种恶意电话的语音特征库的配置方法,位于用户设备侧执行,其特征在于,包括:
    在接听电话过程中,根据用户指令,在检测到用户选择恶意电话选项时,将对方来电标识为恶意电话,并对恶意电话的语音进行一定时长的录音;
    将恶意电话的语音发送给云端服务器,以使所述云端服务器对恶意电话的语音进行语音分析,得到恶意电话的语音特征点,将所述恶意电话的语音特征点保存到恶意电话语音特征数据库中。
  2. 如权利要求1所述的方法,其特征在于,所述恶意电话包括中介电话、推销电话、诈骗电话、骚扰电话或自定义电话;
    在接听电话过程中,根据用户指令,在检测到用户选择恶意电话选项时,将对方来电标识为恶意电话,包括:
    在接听电话过程中,在通话界面上显示电话标签选项,所述电话标签选项中包括中介电话的标签、推销电话的标签、诈骗电话的标签、骚扰电话的标签或自定义电话标签的选项,以便用户根据对方语音在电话标签选项中选择任一标签时,将对方来电标识为用户选择的标签所对应的恶意电话。
  3. 如权利要求2所述的方法,其特征在于,对恶意电话的语音进行一定时长的录音之后,还包括:
    根据用户选择的标签,在录制的恶意电话的语音中添加所述标签信息。
  4. 如权利要求1所述的方法,其特征在于,对恶意电话的语音进行一定时长的录音之后,还包括:
    在录制的恶意电话的语音中添加来电号码。
  5. 如权利要求1-4中任一项所述的方法,其特征在于,
    在接听电话过程中,采集一定时长的对方语音;
    将对方语音发送给云端服务器,以使云端服务器将对方语音进行语音分析,得到对方语音的语音特征点,将对方语音的语音特征点与恶意电话语音特征数据库中的语音特征点进行匹配,在匹配成功时,返回恶意电话的提示;
    根据云端服务器反馈的恶意电话的提示,在通话界面上显示所述恶意电话的提示。
  6. 一种恶意电话的语音特征库的配置方法,位于云端服务器侧执行,其特征在于,包括:
    云端服务器接收用户设备发送的恶意电话的语音,所述恶意电话的语音为用户设备在接听电话过程中,根据用户指令,在检测到用户选择恶意电话选项时,将对方来电标识为恶意电话,并对恶意电话的语音进行一定时长的录音;
    对恶意电话的语音进行语音分析,得到恶意电话的语音特征点;
    将所述恶意电话的语音特征点保存到恶意电话语音特征数据库中。
  7. 如权利要求6所述的方法,其特征在于,若所述恶意电话的语音中携带有所述恶意电话的标签;所述恶意电话的标签包括中介电话的标签、推销电话的标签、诈骗电话的标签、骚扰电话的标签或自定义电话标签;
    则对恶意电话的语音进行语音分析,得到恶意电话的语音特征点之后,包括:
    根据所述恶意电话的语音中携带的标签,建立所述恶意电话的语音特征点与所述标签的对应关系;
    将所述恶意电话的语音特征点与所述标签的对应关系保存到恶意电话语音特征数据库。
  8. 如权利要求6所述的方法,其特征在于,若所述恶意电话的语音中携带有所述恶意电话的来电号码;
    则对恶意电话的语音进行语音分析,得到恶意电话的语音特征点之后,包括:
    根据所述恶意电话的语音中携带的来电号码,建立所述恶意电话的语音特征点与所述来电号码的对应关系;
    将所述恶意电话的语音特征点与所述来电号码的对应关系保存到恶意电话语音特征数据库。
  9. 如权利要求6所述的方法,其特征在于,对恶意电话的语音进行语音分析,得到恶意电话的语音特征点,包括:
    获取所述恶意电话的语音的功率谱;
    将所获得的功率谱通过滤波处理,获得含噪语音的短时功率谱;
    根据所获得的含噪语音的短时功率谱,采用帧平均的方式求取含噪语音的中等时长功率谱;
    对所获得的含噪语音的中等时长功率谱进行不对称滤波和掩蔽抗噪,以得到纯净语音的中等时长功率谱;
    根据所述纯净语音的中等时长功率谱、含噪语音的中等时长功率谱和含噪语音的短时功率谱获取纯净语音的短时功率谱;
    对纯净语音的短时功率谱进行能量归一化处理,以消除乘性噪音;
    对消除了乘性噪音的纯净语音的短时功率谱进行等响度加重
    对等响度加重后的纯净语音的短时功率谱进行指数非线性操作;
    对进行了指数非线性操作后的纯净语音的短时功率谱进行傅立叶逆变换,以求取倒谱系数;
    对倒谱系数进行均值归一化处理,得到所述恶意电话的语音特征点。
  10. 如权利要求9所述的方法,其特征在于,云端服务器接收用户设备发送的恶意电话的语音之后,还包括:
    将所述恶意电话的语音识别成对应的文字信息;
    建立所述恶意电话的文字信息与所述恶意电话的语音特征点之间的对应关系。
  11. 如权利要求6-10中任一项所述的方法,其特征在于,所述恶意电话语音特征数据库中包括:多个恶意电话的来电号码信息、多个恶意电话的语音特征点信息、多个恶意电话的语音的标签信息、多个恶意电话的语音对应的文字信息;
    所述恶意电话语音特征数据库中还包括:恶意电话的来电号码与恶意电话的语音特征点之间的对应关系、恶意电话的语音特征点与标签的对应关 系、恶意电话的语音特征点与文字信息的对应关系、恶意电话的来电号码与文字信息的对应关系、恶意电话的标签与文字信息的对应关系、恶意电话的来电号码和语音特征点以及标签之间的一一对应关系、恶意电话的来电号码、语音特征点、标签以及文字信息之间的一一对应关系。
  12. 如权利要求11所述的方法,其特征在于,还包括:
    云端服务器接收用户设备在接听电话过程中采集的一定时长的对方语音;
    将对方语音进行语音分析,得到对方语音的语音特征点;
    将对方语音的语音特征点与恶意电话语音特征数据库中的语音特征点进行匹配;
    在匹配成功时,向用户设备返回恶意电话的提示以使用户设备根据所述恶意电话的提示,在通话界面上显示所述恶意电话的提示。
  13. 一种恶意电话的语音特征库的配置装置,位于用户设备侧,其特征在于,包括:
    检测模块,用于在接听电话过程中,根据用户指令,检测到用户选择恶意电话选项,将对方来电标识为恶意电话,;
    采集模块,用于对恶意电话的语音进行一定时长的录音;
    发送模块,用于将恶意电话的语音发送给云端服务器,以使所述云端服务器对恶意电话的语音进行语音分析,得到恶意电话的语音特征点,将所述恶意电话的语音特征点保存到恶意电话语音特征数据库中。
  14. 如权利要求13所述的装置,其特征在于,所述恶意电话包括中介电话、推销电话、诈骗电话、骚扰电话或自定义电话;
    所述检测模块具体用于:
    在接听电话过程中,在通话界面上显示电话标签选项,所述电话标签选项中包括中介电话的标签、推销电话的标签、诈骗电话的标签、骚扰电话的标签或自定义电话标签的选项,以便用户根据对方语音在电话标签选项中选择任一标签时,将对方来电标识为用户选择的标签所对应的恶意电话。
  15. 如权利要求14所述的装置,其特征在于,还包括:
    添加模块,用于根据用户选择的标签,在录制的恶意电话的语音中添加所述标签信息。
  16. 如权利要求13所述的装置,其特征在于,
    所述添加模块,还用于在录制的恶意电话的语音中添加来电号码。
  17. 如权利要求13-16中任一项所述的装置,其特征在于,
    所述采集模块,还用于在接听电话过程中,采集一定时长的对方语音;
    所述发送模块,还用于将对方语音发送给云端服务器,以使云端服务器将对方语音进行语音分析,得到对方语音的语音特征点,将对方语音的语音特征点与恶意电话语音特征数据库中的语音特征点进行匹配,在匹配成功时,返回恶意电话的提示;
    显示模块,用于根据云端服务器反馈的恶意电话的提示,在通话界面上显示所述恶意电话的提示。
  18. 一种恶意电话的语音特征库的配置装置,位于云端服务器侧,其特征在于,包括:
    接收模块,用于接收用户设备发送的恶意电话的语音,所述恶意电话的语音为用户设备在接听电话过程中,根据用户指令,在检测到用户选择恶意电话选项时,将对方来电标识为恶意电话,并对恶意电话的语音进行一定时长的录音;
    语音分析模块,用于对恶意电话的语音进行语音分析,得到恶意电话的语音特征点;
    保存模块,用于将所述恶意电话的语音特征点保存到恶意电话语音特征数据库中。
  19. 如权利要求18所述的装置,其特征在于,若所述恶意电话的语音中携带有所述恶意电话的标签;所述恶意电话的标签包括中介电话的标签、推销电话的标签、诈骗电话的标签、骚扰电话的标签或自定义电话标签;
    则所述装置还包括:
    建立模块,用于根据所述恶意电话的语音中携带的标签,建立所述恶意电话的语音特征点与所述标签的对应关系;
    所述保存模块,还用于将所述恶意电话的语音特征点与所述标签的对应关系保存到恶意电话语音特征数据库。
  20. 如权利要求18所述的装置,其特征在于,若所述恶意电话的语音中携带有所述恶意电话的来电号码;
    所述建立模块,还用于根据所述恶意电话的语音中携带的来电号码,建立所述恶意电话的语音特征点与所述来电号码的对应关系;
    所述保存模块,还用于将所述恶意电话的语音特征点与所述来电号码的对应关系保存到恶意电话语音特征数据库。
  21. 如权利要求18所述的装置,其特征在于,所述语音分析模块具体用于:
    获取所述恶意电话的语音的功率谱;
    将所获得的功率谱通过滤波处理,获得含噪语音的短时功率谱;
    根据所获得的含噪语音的短时功率谱,采用帧平均的方式求取含噪语音的中等时长功率谱;
    对所获得的含噪语音的中等时长功率谱进行不对称滤波和掩蔽抗噪,以得到纯净语音的中等时长功率谱;
    根据所述纯净语音的中等时长功率谱、含噪语音的中等时长功率谱和含噪语音的短时功率谱获取纯净语音的短时功率谱;
    对纯净语音的短时功率谱进行能量归一化处理,以消除乘性噪音;
    对消除了乘性噪音的纯净语音的短时功率谱进行等响度加重
    对等响度加重后的纯净语音的短时功率谱进行指数非线性操作;
    对进行了指数非线性操作后的纯净语音的短时功率谱进行傅立叶逆变换,以求取倒谱系数;
    对倒谱系数进行均值归一化处理,得到所述恶意电话的语音特征点。
  22. 如权利要求21所述的装置,其特征在于,
    所述语音分析模块,还用于将所述恶意电话的语音识别成对应的文字信息;
    所述建立模块,还用于建立所述恶意电话的文字信息与所述恶意电话的语音特征点之间的对应关系。
  23. 如权利要求18-22中任一项所述的装置,其特征在于,所述恶意电话语音特征数据库中包括:多个恶意电话的来电号码信息、多个恶意电话的语音特征点信息、多个恶意电话的语音的标签信息、多个恶意电话的语音对应的文字信息;
    所述恶意电话语音特征数据库中还包括:恶意电话的来电号码与恶意电话的语音特征点之间的对应关系、恶意电话的语音特征点与标签的对应关系、恶意电话的语音特征点与文字信息的对应关系、恶意电话的来电号码与文字信息的对应关系、恶意电话的标签与文字信息的对应关系、恶意电话的来电号码和语音特征点以及标签之间的一一对应关系、恶意电话的来电号码、语音特征点、标签以及文字信息之间的一一对应关系。
  24. 如权利要求23所述的装置,其特征在于,
    所述接收模块,还用于接收用户设备在接听电话过程中采集的一定时长的对方语音;
    所述语音分析模块,还用于将对方语音进行语音分析,得到对方语音的语音特征点;
    匹配模块,用于将对方语音的语音特征点与恶意电话语音特征数据库中的语音特征点进行匹配;
    所述发送模块,用于在匹配成功时,向用户设备返回恶意电话的提示以使用户设备根据所述恶意电话的提示,在通话界面上显示所述恶意电话的提示。
  25. 一种计算机程序,包括计算机可读代码,当所述计算机可读代码在计算设备上运行时,导致所述计算设备执行根据权利要求1-12中的任一项所述的恶意电话的语音特征库的配置方法。
  26. 一种计算机可读介质,其中存储了如权利要求25所述的计算机程序。
PCT/CN2016/106150 2015-11-20 2016-11-16 恶意电话的语音特征库的配置方法和装置 WO2017084592A1 (zh)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN201510808778.XA CN106790949A (zh) 2015-11-20 2015-11-20 恶意电话的语音特征库的配置方法和装置
CN201510808778.X 2015-11-20

Publications (1)

Publication Number Publication Date
WO2017084592A1 true WO2017084592A1 (zh) 2017-05-26

Family

ID=58717317

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2016/106150 WO2017084592A1 (zh) 2015-11-20 2016-11-16 恶意电话的语音特征库的配置方法和装置

Country Status (2)

Country Link
CN (1) CN106790949A (zh)
WO (1) WO2017084592A1 (zh)

Families Citing this family (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN108234485B (zh) * 2017-12-30 2020-09-01 广东世纪网通信设备股份有限公司 基于voip平台的诈骗声纹获取装置及利用该装置来拦截诈骗电话的方法、装置和系统
CN108848507A (zh) * 2018-05-31 2018-11-20 厦门快商通信息技术有限公司 一种不良电信用户信息采集方法
CN109547621A (zh) * 2018-12-11 2019-03-29 平安科技(深圳)有限公司 一种信息提醒方法、装置、电子设备及存储介质
CN110233943B (zh) * 2019-06-18 2021-07-09 深圳市珍爱云信息技术有限公司 获取号码标记的方法、装置、计算机设备和存储介质
CN110619893A (zh) * 2019-09-02 2019-12-27 合肥工业大学 语音信号的时频特征提取及人工智能情绪监测方法

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US8396711B2 (en) * 2006-05-01 2013-03-12 Microsoft Corporation Voice authentication system and method
CN103731832A (zh) * 2013-12-26 2014-04-16 黄伟 防电话、短信诈骗的系统和方法
CN104410973A (zh) * 2014-11-20 2015-03-11 北京新讯世纪信息技术有限公司 一种播放录音的诈骗电话识别方法和系统
CN104751848A (zh) * 2013-12-25 2015-07-01 三亚中兴软件有限责任公司 通话声音识别方法及装置
CN105007361A (zh) * 2015-06-13 2015-10-28 安徽味唯网络科技有限公司 一种智能自动屏蔽骚扰电话的方法

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US8396711B2 (en) * 2006-05-01 2013-03-12 Microsoft Corporation Voice authentication system and method
CN104751848A (zh) * 2013-12-25 2015-07-01 三亚中兴软件有限责任公司 通话声音识别方法及装置
CN103731832A (zh) * 2013-12-26 2014-04-16 黄伟 防电话、短信诈骗的系统和方法
CN104410973A (zh) * 2014-11-20 2015-03-11 北京新讯世纪信息技术有限公司 一种播放录音的诈骗电话识别方法和系统
CN105007361A (zh) * 2015-06-13 2015-10-28 安徽味唯网络科技有限公司 一种智能自动屏蔽骚扰电话的方法

Also Published As

Publication number Publication date
CN106790949A (zh) 2017-05-31

Similar Documents

Publication Publication Date Title
WO2017084592A1 (zh) 恶意电话的语音特征库的配置方法和装置
WO2017084591A1 (zh) 恶意电话的识别方法和装置
US11140267B1 (en) Voice and speech recognition for call center feedback and quality assurance
CN104966053B (zh) 人脸识别方法及识别系统
WO2017076314A1 (zh) 自适应识别骚扰电话的处理方法及系统
US9860372B2 (en) Device and method for displaying information relevant to unknown incoming/outgoing phone number
US20190373105A1 (en) Cognitive telephone fraud detection
CN103888581B (zh) 一种通信终端及其记录通话信息的方法
CN105141787A (zh) 服务录音的合规检查方法及装置
WO2014069120A1 (ja) 分析対象決定装置及び分析対象決定方法
CN107808423A (zh) 一种基于服务机器人的员工考勤签到系统与实现方法
US9552512B2 (en) Personalized, automated receptionist
US9716789B2 (en) Method, device, terminal and server for replying call
RU2008152794A (ru) Идентификация медиаданных
CN103916513A (zh) 在通信终端记录通话信息的方法和设备
US20150066925A1 (en) Method and Apparatus for Classifying Data Items Based on Sound Tags
US11341956B1 (en) Electronic signatures via voice for virtual assistants' interactions
WO2017059679A1 (zh) 一种帐号处理方法及装置
US20160351185A1 (en) Voice recognition device and method
JP2016053598A (ja) 通話装置、声紋登録方法およびプログラム
WO2016107277A1 (zh) 一种基于电话号码的信息加载方法及装置
WO2020226994A1 (en) Combination of real-time analytics and automation
TW201626364A (zh) 自動還原丟失語音資訊的系統與方法
CN104580638B (zh) 一种电话号码标注的方法及装置
KR101033870B1 (ko) 스팸 전화 처리 방법 및 장치

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: 16865769

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: 16865769

Country of ref document: EP

Kind code of ref document: A1