TW202001643A - System for determining identity through voiceprint and voice password and method thereof - Google Patents

Abstract

A system for determining identity through a voiceprint and a voice password and a method thereof are provided. By receiving a voice, and determining whether the voice is verified in accordance with a voiceprint of the voice and message of the voice, the system and the method can confirm recognized voice made by real people, and can achieve the effect of increasing security of identity recognized.

Description

System and method for judging identity through voiceprint and voice password

A body identification system and method thereof, particularly a system and method for judging identity through voiceprint and voice password.

Voiceprint is the sound wave spectrum carrying language information displayed by the instrument. It can be learned from modern scientific research that no matter whether the speaker whispered softly or intentionally imitated the voice and tone of others, even if imitated brilliantly, the voiceprint is always different, that is to say, the vocal organ tongue used by people when speaking , Teeth, throat, lungs, nasal cavity are different in size and shape, so any two people have different voiceprints, so the voiceprint is unique. In addition, after adulthood, the sound can remain unchanged for a long time, which means that the voiceprint also has relative stability.

According to different application scenarios, the current voiceprint recognition can be divided into speaker identification (Speaker Identification, SI) and speaker verification (Speaker Verification, SV). Among them, speaker confirmation refers to judging whether a piece of speech comes from a specific user, that is to say, through voiceprint recognition technology, the identity of the speaker can be judged from the speech of the speaker. However, although the identity of the speaker can be determined through voiceprint recognition, and others cannot imitate it, voiceprint recognition cannot distinguish whether the voice was issued by a real person or played by a sound playback device, therefore, interested persons may record The voice of the speaker is used to obtain the voiceprint of the speaker, so as to forge the voice of the speaker, so as to embezzle the identity of the speaker.

In summary, it can be seen that the voiceprint recognition technology has been unable to determine whether the voice is uttered by a real person for a long time in the prior art. Therefore, it is necessary to propose improved technical means to solve this problem.

In view of the problem in the prior art that the voiceprint recognition technology cannot determine whether the voice is uttered by a real person, the present invention discloses a system and method for determining identity through voiceprint and voice password, in which:

The system for judging identity by voiceprint and voice password disclosed by the present invention at least includes: a voice receiving module for receiving the first voice signal and the second voice signal; a voiceprint recognition module for judging the first voice signal Whether the voiceprint matches the pre-established voiceprint; the voice recognition module is used to determine whether the first voice signal is consistent with the predetermined password, and to determine whether the second voice signal is consistent with the voice password; the prompt module is used to When the voiceprint of the first voice signal matches the pre-established voiceprint, and the first voice signal matches the predetermined password, prompt to input the second voice signal; the operation execution module is used to match the second voice signal and the voice password When the judgment result is verified, the operation corresponding to the first voice signal is performed.

The method for judging identity by voiceprint and voice password disclosed in the present invention includes at least the steps of: receiving a first voice signal; judging that the voiceprint of the first voice signal matches the pre-established voiceprint, and the first voice signal is specific When the passwords match, prompt to input the second voice signal; receive the second voice signal; when it is judged that the second voice signal matches the voice password, the judgment result is verified, and the operation corresponding to the first voice signal is performed.

The system and method disclosed by the present invention are as above. The difference between the present invention and the prior art is that the present invention solves the problems of the prior art by determining whether the result of the voice signal and the content of the voice signal are verified after receiving the voice signal. Problem, and can achieve the technical effect of increasing the security of identity recognition.

The following will describe the features and implementations of the present invention in detail with reference to the drawings and examples. The content is sufficient for any person skilled in the relevant arts to easily fully understand and implement the technical means applied to solve the technical problems of the present invention and implement accordingly, thereby realizing The achievable effect of the invention.

In this case, voiceprint recognition and voice recognition can be performed on the voice signal input by the user, so as to determine the user's identity through voiceprint recognition, and confirm the user as the user through voice recognition, so as to prevent the user's voiceprint from being forged or misappropriated.

The following describes the system operation of the present invention with the "Figure 1" system architecture diagram proposed by the present invention for determining identity through voiceprint and voice password. As shown in "Figure 1", the system of the present invention includes a voice receiving module 110, a voiceprint recognition module 131, a voice recognition module 132, a prompt module 150, an operation execution module 160, and an additional emotion evaluation Module 133, warning module 170, authority judgment module 180, and password generation module 190.

The voice receiving module 110 is responsible for receiving the first voice signal and also for receiving the second voice signal, wherein the first voice signal and the second voice signal are usually different voice signals that are input at different times. In some embodiments, the voice receiving module 110 may also receive the third voice signal.

Generally speaking, the voice receiving module 110 can receive the first voice signal, the second voice signal, the third voice signal, and the set voice signal from an external microphone or other voice input device (not shown), but the present invention does not This is the limit.

The voiceprint recognition module 131 is responsible for determining whether the voiceprint of the first voice signal received by the voice receiving module 110 matches the pre-established target voiceprint. The voiceprint recognition module 131 can usually establish a target voiceprint through the voice receiving module 110, but the invention is not limited thereto.

The voice recognition module 132 is responsible for judging whether the first voice signal received by the voice receiving module 110 is consistent with a predefined password. Wherein, the predetermined password is usually a pre-recorded voice signal, and the voice recognition module 132 can convert the predetermined password into corresponding password text, and can convert the first voice signal received by the voice receiving module 110 into corresponding voice text , And compare whether the voice text and password text generated by the conversion are the same, so as to determine whether the first voice signal is consistent with the predetermined password, but the present invention is not limited to this, for example, the predetermined password may also be text data, voice recognition mode The group 132 may compare and convert the first voice signal to generate voice text and a predetermined password to determine whether the first voice signal and the predetermined password are consistent.

The voice recognition module 132 is also responsible for determining whether the second voice signal received by the voice receiving module 110 matches the voice password. Generally speaking, the voice password is usually text data, and the voice recognition module 132 can convert the second voice signal into the corresponding voice text, and compare whether the converted voice text and the voice password are the same, but the present invention does not use this Limited.

The emotion evaluation module 133 may determine the emotional state according to the voiceprint of the first voice signal and/or the second voice signal received by the voice receiving module 110. The emotion state judged by the emotion assessment module 133 may be normal, happy, angry, sad, etc., indicating normal emotional states, or may be abnormal emotional states such as fear or fear, but the emotional states mentioned in the present invention are not based on the above Limited.

The emotion evaluation module 133 can also confirm the emotional state according to the third voice signal received by the voice receiving module 110, that is, determine whether the emotional state is abnormal again according to the third voice signal, if the emotional state determined this time still indicates Abnormal, the emotion evaluation module 133 can confirm that the emotional state is abnormal, and if the emotional state determined this time indicates normal, the operation execution module 160 can execute normally. The emotion evaluation module 133 can determine whether the emotional state is abnormal according to the voiceprint of the third voice signal, and can also determine whether the emotional state is abnormal according to the content of the third voice signal.

The prompt module 150 is responsible for prompting for the second input when the voiceprint recognition module 131 judges that the voiceprint of the first voice signal matches the pre-established target voiceprint, and the voice recognition module 132 judges that the first voice signal matches the predetermined password Voice signal. The prompt module 150 can play a pre-created voice message or display a pre-created text message to prompt for the input of a second voice signal, but the invention is not limited to this.

The prompt module 150 may also generate an abnormal question when the emotion evaluation module 133 determines that the emotional state indicates abnormality, and prompt to input a third voice signal corresponding to the abnormal question. The abnormal question mentioned in the present invention is usually a predetermined set of questions, where each question can correspond to a correct answer, and the emotion evaluation module 133 can judge that the emotional state is normal when the third voice signal matches the correct answer, Furthermore, when the third voice signal does not match the correct answer, it can be judged that the emotional state is abnormal; in some embodiments, each question can also have multiple corresponding answers, for example, a solution indicating that the emotional state is normal and an emotional state. Abnormal answers, etc. When the third voice signal matches the normal answer, the emotion assessment module 133 can determine that the emotional state is normal, and when the third voice signal matches the abnormal answer, the emotion assessment module 133 can determine that the emotional state is abnormal .

The operation execution module 160 is responsible for when the voice recognition module 132 judges that the second voice signal is consistent with the voice password, the judgment result is verified, and performs an operation corresponding to the first voice signal, such as executing a specific function. If the voice recognition module 132 determines that the second voice signal does not match the voice password, the operation execution module 160 may determine that the result has not passed the verification and does not perform the operation corresponding to the first voice signal.

The warning module 170 can warn when the emotion evaluation module 133 judges that the emotional state is abnormal, or it can choose whether to warn according to the third voice signal received by the voice receiving module 110. Wherein, the warning module 170 can perform a warning operation to perform warning, such as an alarm, an audible alarm, etc., but the invention is not limited to this.

The authority judging module 180 can judge whether the operation corresponding to the first voice signal can be executed by the operation execution module 160 according to the voiceprint of the first voice signal received by the voice receiving module 110. In more detail, the permission judgment module 180 can determine the user's identity according to the voiceprint of the first voice signal, and read the permission corresponding to the determined user's identity, thereby determining whether the user's identity has the ability to execute with the first The authority of the operation corresponding to the voice signal is to determine whether the operation corresponding to the first voice signal can be performed according to the read authority.

The password generation module 190 can generate the voice password required by the voice recognition module 132, and can transmit the generated voice password to a designated device (not shown). The designated device mentioned in the present invention may be an email server, a mobile phone, etc., but the present invention is not limited to this. Generally speaking, the password generation module 190 can randomly generate a voice password, or can generate a voice password according to a specific rule, and the present invention is not particularly limited.

Next, an embodiment is used to explain the operation system and method of the present invention, and please refer to the "Figure 2A" flowchart of the method for judging identity by voiceprint and voice password proposed by the present invention. In this embodiment, it is assumed that the present invention is applied to electronic devices such as audio-visual players, speakers, mobile phones, tablets, screens, and safes.

First, the electronic device can store the owner's voiceprint. In this embodiment, the voice receiving module 110 can receive the voice signal sent by the owner, and the voiceprint recognition module 131 can obtain the voiceprint of the owner according to the voice signal received by the voice receiving module 110, and then The obtained voiceprint is used as the target voiceprint.

Afterwards, when a user wants to use the electronic device, the voice receiving module 110 can receive the first voice signal sent by the user regardless of whether the user owns it or not (step 210). In this embodiment, it is assumed that the voice receiving module 110 is connected to the microphone on the electronic device, and the voice receiving module 110 continues to receive the voice signal, so as to receive the first voice signal input by the user to the microphone.

After the voice receiving module 110 receives the first voice signal (step 210), the voiceprint recognition module 131 can determine whether the voiceprint of the first voice signal received by the voice receiving module 110 is in accordance with the target previously established by the owner When the voiceprint matches (step 220), the voice recognition module 132 can also determine whether the first voice signal received by the voice receiving module 110 matches a specific password (step 230). In practice, the voiceprint recognition module 131 determines whether the voiceprint of the first voice signal matches the target voiceprint previously established by the owner (step 220) and the voice recognition module 132 determines whether the first voice signal matches the specific password ( Step 230) There is no order relationship, that is to say, in the present invention, the voice recognition module 132 can first determine whether the first voice signal matches a specific password (step 230), and then the voiceprint recognition module 131 Determine whether the voiceprint of the first voice signal matches the target voiceprint previously established by the owner (step 220).

If the voiceprint recognition module 131 determines that the voiceprint of the first voice signal does not match the target voiceprint previously established by the owner, or if the voice recognition module 132 determines that the first voice signal does not match the specific password, the operation execution module 160 It can be judged that the result has not passed the verification, and the operation execution module 160 will not perform the operation corresponding to the first voice signal, for example, unlock the electronic device, so that the user cannot use the electronic device. If the voiceprint recognition module 131 determines that the voiceprint of the first voice signal matches the target voiceprint previously established by the owner, and the voice recognition module 132 determines that the first voice signal matches the specific password, the prompt module 150 may prompt The second voice signal is input (step 241), and the voice receiving module 110 can receive the second voice signal (step 245). In this embodiment, it is assumed that the prompt module 150 can play a prompt message prompting the user to input a preset voice password, thereby prompting the user to say the voice password, so that the voice receiving module 110 can receive the second voice signal.

After the voice receiving module 110 receives the second voice signal from the user (step 245), the voice recognition module 132 can determine whether the second voice signal received by the voice receiving module 110 matches the voice password (step 250 ). If the voice recognition module 132 judges that the second voice signal matches the voice password, the operation execution module 160 can judge that the result passes the verification and can perform the operation corresponding to the first voice signal (step 295), for example, to release the electronic device Locked, so that the user can use the electronic device; and if the voice recognition module 132 determines that the second voice signal does not match the voice password, the operation execution module 160 can determine that the result has not passed the verification, and the operation execution module 160 will not The operation corresponding to the first voice signal is performed, that is, the electronic device is unlocked, so the user cannot use the electronic device.

In this way, after judging the identity of the user through the voiceprint, the present invention additionally checks whether the user sends out the second input signal of the voice password to avoid the situation that the user can use the electronic device by embezzling the owner's voiceprint.

In the above embodiment, if the electronic device includes the password generation module 190, before or after the prompt module 150 prompts to input the second voice signal (step 241), the voice receiving module 110 receives the second voice signal (step 245) The password generation module 190 can dynamically generate a voice password and transmit the generated voice password to a designated device. In this embodiment, assuming that the designated device is the owner's mobile phone, the password generation module 190 can randomly generate a voice password, and use the generated voice password to send the password to the text message, e-mail, instant messaging, near-end communication, etc. The generated voice password is transmitted to the owner's mobile phone, so that when the user is the owner, the user can obtain the voice password.

In addition, in the above embodiment, if the electronic device includes the emotion evaluation module 133, as shown in the flow of "Figure 2B", after the voice recognition module 132 determines that the second voice signal matches the voice password, the operation execution mode Before the judgment result of the group 160 is verified and the operation corresponding to the first voice signal is performed (step 295), the emotion evaluation module 133 may judge according to the first voice signal and/or the second voice signal received by the voice receiving module 110 Whether the user's emotional state is abnormal (step 260), if the emotional evaluation module 133 determines that the user's emotional state is normal, the operation execution module 160 can determine that the result is verified and can perform the operation corresponding to the first voice signal ( Step 295) In this embodiment, that is, the locked state of the electronic device is released.

If the emotion assessment module 133 determines that the user's emotional state is abnormal, the prompt module 150 can generate an abnormal question (step 271), and can provide the generated abnormal question to the user. At the same time, the voice receiving module 110 can Prepare to receive the third voice signal generated by the user answering the abnormal question (step 275). In this embodiment, assuming that the abnormal question is a question pre-recorded by the owner, the prompt module 150 can play back the question pre-recorded by the owner, and the user can speak the answer to the question played by the prompt module 150 to make the voice received The module 110 can receive the response sent by the user, that is, the third voice signal.

After the voice receiving module 110 receives the third voice signal (step 275), the emotion evaluation module 133 can confirm whether the user's emotional state is abnormal according to the third voice signal received by the voice receiving module 110 (step 281) If, after confirmation, the emotion assessment module 133 determines that the user's emotional state is normal, the operation execution module 160 can determine that the result passes the verification and can perform the operation corresponding to the first voice signal (step 295). In this implementation In the example, that is, the locked state of the electronic device is released.

If the emotion evaluation module 133 still judges that the user's emotional state is abnormal after confirmation, the warning module 170 can select whether to warn according to the content of the third voice signal received by the voice receiving module 110 (step 285). In this embodiment, it is assumed that the abnormal question is a predetermined set of questions, where each question has a solution indicating that the emotional state is normal and a solution indicating that the emotional state is abnormal, if the third voice signal matches the solution indicating that the emotional state is normal , The warning module 170 can choose not to warn, or the third voice signal does not match the answer indicating that the emotional state is normal, but the voiceprint of the third voice signal indicates that the emotional state is normal, then the warning module 170 can choose not to warn; and if The third voice signal is consistent with the answer indicating that the emotional state is abnormal. The warning module 170 may choose to warn, and the warning module 170 may issue an alarm, or may notify the administrator or the police station.

In summary, it can be seen that the difference between the present invention and the prior art lies in the technical means of judging whether the result of verification is based on the voiceprint and content of the voice signal after receiving the voice signal. Solve the problem that the voiceprint recognition technology in the prior art cannot determine whether the voice is uttered by a real person, and then achieve the technical effect of increasing the security of identity recognition.

Furthermore, the method for determining the identity by voiceprint and voice password of the present invention can be implemented in hardware, software, or a combination of hardware and software, and can also be implemented in a centralized manner in a computer system or dispersed in several components with different components Decentralized implementation of interconnected computer systems.

Although the disclosed embodiments of the present invention are as described above, the content described is not intended to directly limit the patent protection scope of the present invention. Anyone who has ordinary knowledge in the technical field to which the present invention belongs, without making any departure from the spirit and scope disclosed by the present invention, makes slight modifications to the form and details of the implementation of the present invention, all belong to the patent protection of the present invention range. The scope of patent protection of the present invention shall still be determined by the scope of the attached patent application.

110‧‧‧Voice receiving module 131‧‧‧ voiceprint recognition module 132‧‧‧Speech recognition module 133‧‧‧Emotion Assessment Module 150‧‧‧ prompt module 160‧‧‧Operation execution module 170‧‧‧Alarm module 180‧‧‧authority judgment module 190‧‧‧ Password generation module Step 210: Receive the first voice signal Step 220: Determine whether the voiceprint of the first voice signal matches the target voiceprint Step 230: Determine whether the first voice signal matches the specific password Step 241‧‧‧Prompt to input the second voice signal Step 245‧‧‧ Receive the second voice signal Step 250: Determine whether the second voice signal matches the voice password Step 260‧‧‧ judge whether the emotional state is abnormal Step 271‧‧‧Abnormal question Step 275‧‧‧ Receive the third voice signal Step 281‧‧‧Confirm whether the emotional state is abnormal according to the voiceprint of the third voice signal Step 285‧‧‧Select whether to warn based on the content of the third voice signal Step 295‧‧‧ Perform the operation corresponding to the first voice signal

FIG. 1 is a system architecture diagram of determining identity by voiceprint and voice password according to the present invention. FIG. 2A is a flowchart of a method for judging identity through voiceprint and voice password according to the present invention. FIG. 2B is a flowchart of a method for selecting and warning according to the emotional state of a voice signal according to the present invention.

Step 210: Receive the first voice signal

Step 220: Determine whether the voiceprint of the first voice signal matches the target voiceprint

Step 230: Determine whether the first voice signal matches the specific password

Step 241‧‧‧Prompt to input the second voice signal

Step 245‧‧‧ Receive the second voice signal

Step 250: Determine whether the second voice signal matches the voice password

Step 295‧‧‧ Perform the operation corresponding to the first voice signal

Claims (10)

A method for judging identity by voiceprint and voice password, the method includes at least the following steps: receiving a first voice signal; judging that the voiceprint of the first voice signal matches the pre-established target voiceprint, and the first voice signal When it matches a specific password, it prompts to enter a second voice signal; receives the second voice signal; and when it is determined that the second voice signal matches a voice password, performs an operation corresponding to the first voice signal. The method for determining identity by voiceprint and voice password as described in item 1 of the patent application scope, wherein the method further includes generating the voice password before the step of receiving the second voice signal, and sending the voice password to a designated device Steps. The method for determining identity by voiceprint and voice password as described in item 1 of the patent application scope, wherein the method further includes the step of performing the operation corresponding to the first voice signal based on the first voice signal and/or The voiceprint of the second voice signal determines an emotional state, and a step of warning when the emotional state indicates abnormality. The method for judging identity through voiceprint and voice password as described in Item 3 of the patent application scope, wherein the method of warning when the emotional state indicates abnormality, further includes generating an abnormal question and receiving corresponding to the abnormal question A third voice signal, and when it is confirmed that the emotional state indicates abnormality according to the third voice signal, the step of whether to warn is selected according to the third voice signal. The method for judging identity by voiceprint and voice password as described in item 1 of the patent scope, wherein the method further includes judging the voiceprint according to the first voice signal before performing the step corresponding to the operation of the first voice signal Whether the steps of the operation corresponding to the first voice signal can be performed. A system for judging identity through voiceprint and voice password, the system at least includes: a voice receiving module for receiving a first voice signal and a second voice signal; a voiceprint identification module for To determine whether the voiceprint of the first voice signal matches the pre-established target voiceprint; a voice recognition module to determine whether the first voice signal matches a predetermined password and to determine the second voice signal Whether it matches a voice password; a prompt module for the voiceprint recognition module to judge that the voiceprint of the first voice signal matches the pre-established voiceprint, and the voice recognition module judges the first voice signal When the predetermined password matches, the second voice signal is prompted for input; and an operation execution module is used to perform an operation corresponding to the first voice signal when the second voice signal matches the voice password. As described in Item 6 of the patent application scope, a system for determining identity by voiceprint and voice password, wherein the system further includes a password generation module for generating the voice password and transmitting the voice password to a designated device. The system for determining identity by voiceprint and voice password as described in item 6 of the patent application scope, wherein the system further includes an emotion evaluation module and a warning module, the emotion evaluation module is used to determine the first voice signal And/or the voiceprint of the second voice signal determines an emotional state, and the warning module is used to warn when the emotional state is abnormal. The system for determining identity through voiceprint and voice password as described in item 8 of the patent application scope, wherein the prompt module is further used to generate an abnormal question when the emotional state indicates abnormality, and the voice receiving module is further used to receive A third voice signal corresponding to the abnormal question, so that the emotion evaluation module confirms the emotional state according to the third voice signal, and the warning module is used for confirming that the emotional state is abnormal The third voice signal selects whether to warn. The system for determining identity by voiceprint and voice password as described in item 6 of the patent application scope, wherein the system further includes a permission judgment module for judging whether or not the first voice signal can be executed according to the voiceprint of the first voice signal Operation corresponding to a voice signal.

Images