Method, device, electronic equipment and computer readable storage medium for echo detection

Abstract

This application belongs to the field of speech processing technology and discloses a method, apparatus, electronic device, and computer-readable storage medium for echo detection. The method includes: acquiring a first speech sent to a target device; acquiring a second speech responded by the target device based on the first speech; determining the similarity between the first speech and the second speech; and obtaining an echo detection result based on the similarity. This allows for accurate detection of speech echoes, thereby avoiding voice interaction errors caused by echoes.

Description

Technical Field

[0001] This application relates to the field of speech processing technology, and more specifically, to methods, apparatus, electronic devices, and computer-readable storage media for echo detection. Background Technology

[0002] In some voice call scenarios (such as when a user is making a hands-free call in an open space), there may be an echo problem.

[0003] For example, User A and User B are having a call using a first terminal and a second terminal, respectively. Because User B is in an open area making a hands-free call, there is an echo. Therefore, the second terminal transmits the echo of User A's voice to User A's first terminal, causing User A to mistake the echo for User B's voice reply, resulting in a call error between the users.

[0004] Therefore, an echo detection technology is needed to detect whether there is an echo in a voice conversation in order to perform echo processing. Summary of the Invention

[0005] The purpose of this application is to provide a method, apparatus, electronic device, and computer-readable storage medium for echo detection, so as to accurately detect voice echo during voice conversations and avoid voice interaction errors caused by echo.

[0006] On the one hand, a method for echo detection is provided, including:

[0007] Acquire the first voice message sent to the target device;

[0008] Acquire the second voice response from the target device based on the first voice response;

[0009] Determine the similarity between the first and second speech;

[0010] Echo detection results are obtained based on similarity.

[0011] In one implementation, determining the similarity between a first speech and a second speech includes:

[0012] Extract the first audio features of the first speech and the second audio features of the second speech;

[0013] Determine the audio feature matching degree between the first audio feature and the second audio feature;

[0014] The degree of audio feature matching is defined as similarity.

[0015] In one implementation, determining the similarity between a first speech and a second speech includes:

[0016] Perform text conversion on the first speech to obtain the first speech text;

[0017] The second speech is converted into text to obtain the second speech text;

[0018] Determine the text feature matching degree between the first and second speech texts;

[0019] The text feature matching degree is defined as similarity.

[0020] In one implementation, determining the similarity between a first speech and a second speech includes:

[0021] Determine the audio feature matching degree between the first and second speech;

[0022] Determine the text feature matching degree between the first and second speech;

[0023] Similarity is determined based on the matching degree of audio features and the matching degree of text features.

[0024] In one implementation, the similarity is determined based on the audio feature matching degree and the text feature matching degree, including:

[0025] The similarity score is obtained by weighted summation of audio feature matching scores and text feature matching scores.

[0026] Alternatively, the similarity score can be determined by combining the audio feature matching score and the text feature matching score.

[0027] In one implementation, obtaining echo detection results based on similarity includes:

[0028] If the similarity is determined to meet the conditions for speech anomalies, then an echo is confirmed to exist.

[0029] Alternatively, if the similarity is determined to meet the conditions for speech anomalies, then an echo anomaly is determined to exist, and the number of echo anomalies is accumulated. If the number of echo anomalies is determined to be higher than the threshold, then an echo is determined to exist.

[0030] In one embodiment, the method further includes:

[0031] If the echo detection results indicate the presence of an echo, then the second speech is filtered.

[0032] On the one hand, an echo detection device is provided, comprising:

[0033] The first acquisition unit is used to acquire the first voice message sent to the target device;

[0034] The second acquisition unit is used to acquire the second voice response from the target device based on the first voice response;

[0035] A determining unit is used to determine the similarity between the first speech and the second speech;

[0036] The acquisition unit is used to obtain echo detection results based on similarity.

[0037] In one implementation, the determining unit is used to:

[0038] Extract the first audio features of the first speech and the second audio features of the second speech;

[0039] Determine the audio feature matching degree between the first audio feature and the second audio feature;

[0040] The degree of audio feature matching is defined as similarity.

[0041] In one implementation, the determining unit is used to:

[0042] Perform text conversion on the first speech to obtain the first speech text;

[0043] The second speech is converted into text to obtain the second speech text;

[0044] Determine the text feature matching degree between the first and second speech texts;

[0045] The text feature matching degree is defined as similarity.

[0046] In one implementation, the determining unit is used to:

[0047] Determine the audio feature matching degree between the first and second speech;

[0048] Determine the text feature matching degree between the first and second speech;

[0049] Similarity is determined based on the matching degree of audio features and the matching degree of text features.

[0050] In one implementation, the determining unit is used to:

[0051] The similarity score is obtained by weighted summation of audio feature matching scores and text feature matching scores.

[0052] Alternatively, the similarity score can be determined by combining the audio feature matching score and the text feature matching score.

[0053] In one embodiment, the obtaining unit is used to:

[0054] If the similarity is determined to meet the conditions for speech anomalies, then an echo is confirmed to exist.

[0055] Alternatively, if the similarity is determined to meet the conditions for speech anomalies, then an echo anomaly is determined to exist, and the number of echo anomalies is accumulated. If the number of echo anomalies is determined to be higher than the threshold, then an echo is determined to exist.

[0056] In one embodiment, the obtaining unit is further configured to:

[0057] If the echo detection results indicate the presence of an echo, then the second speech is filtered.

[0058] On one hand, an electronic device is provided, including a processor and a memory storing computer-readable instructions that, when executed by the processor, perform the steps of the method provided in any of the alternative implementations of echo detection described above.

[0059] On the one hand, a computer-readable storage medium is provided on which a computer program is stored, which, when executed by a processor, performs the steps of the method provided in any of the various alternative implementations of echo detection described above.

[0060] On the one hand, a computer program product is provided that, when run on a computer, causes the computer to perform the steps of the method provided in any of the alternative implementations of echo detection described above.

[0061] The echo detection method, apparatus, electronic device, and computer-readable storage medium provided in this application embodiment acquire a first voice message sent to a target device; acquire a second voice message responded by the target device based on the first voice message; determine the similarity between the first and second voice messages; and obtain an echo detection result based on the similarity. This allows for accurate detection of voice echoes, thus avoiding voice interaction errors caused by echoes.

[0062] Other features and advantages of this application will be set forth in the description which follows, and will be apparent in part from the description, or may be learned by practicing the application. The objectives and other advantages of this application may be realized and obtained by means of the structures particularly pointed out in the written description, claims, and drawings. Attached Figure Description

[0063] To more clearly illustrate the technical solutions of the embodiments of this application, the accompanying drawings used in the embodiments of this application will be briefly introduced below. It should be understood that the following drawings only show some embodiments of this application and should not be regarded as a limitation of the scope. For those skilled in the art, other related drawings can be obtained based on these drawings without creative effort.

[0064] Figure 1 A flowchart of an echo detection method provided in an embodiment of this application;

[0065] Figure 2 A structural block diagram of an echo detection device provided in an embodiment of this application;

[0066] Figure 3 This is a schematic diagram of the structure of an electronic device provided in an embodiment of this application. Detailed Implementation

[0067] The technical solutions of the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only a part of the embodiments of this application, and not all of the embodiments. The components of the embodiments of this application described and shown in the accompanying drawings can generally be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of this application provided in the accompanying drawings is not intended to limit the scope of the claimed application, but merely represents selected embodiments of this application. All other embodiments obtained by those skilled in the art based on the embodiments of this application without inventive effort are within the scope of protection of this application.

[0068] First, some of the terms used in the embodiments of this application will be explained to facilitate understanding by those skilled in the art.

[0069] Terminal devices can be mobile terminals, fixed terminals, or portable terminals, such as mobile phones, sites, units, devices, multimedia computers, multimedia tablets, internet nodes, communicators, desktop computers, laptop computers, notebook computers, netbook computers, tablet computers, personal communication system devices, personal navigation devices, personal digital assistants, audio / video players, digital cameras / camcorders, positioning devices, television receivers, radio broadcast receivers, e-book devices, gaming devices, or any combination thereof, including accessories and peripherals of these devices, or any combination thereof. It is also foreseeable that terminal devices can support any type of user-facing interface (e.g., wearable devices).

[0070] Servers can be independent physical servers, server clusters or distributed systems composed of multiple physical servers, or cloud servers that provide basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communication, middleware services, domain name services, security services, and big data and artificial intelligence platforms.

[0071] In order to accurately detect voice echo during voice conversations and avoid voice interaction problems caused by echo, embodiments of this application provide a method, apparatus, electronic device, and computer-readable storage medium for echo detection.

[0072] This application applies to scenarios where different callers conduct voice calls through different calling devices. The callers can be humans or robots. The calling devices can be servers or terminal devices.

[0073] See Figure 1 The diagram shown is a flowchart of an echo detection method provided in an embodiment of this application. The specific implementation process of this method is as follows:

[0074] Step 100: Obtain the first voice message sent to the target device.

[0075] Specifically, the first caller sends a first voice message to the target device of the second caller through the call device.

[0076] The target device can be a server or a terminal device.

[0077] Step 101: Obtain the second voice response from the target device based on the first voice response.

[0078] It should be noted that, due to the possibility of echo, the second voice message may be returned by the second caller to the first caller's calling device through the target device, or the target device may generate an echo of the first voice message, use the echo as the second voice message, and return it to the first caller's calling device.

[0079] Step 102: Determine the similarity between the first and second speech.

[0080] In one implementation, step 102 can be performed in any of the following ways:

[0081] Method 1: Determine the similarity based on the matching degree of audio features between the first and second speech.

[0082] Specifically, the first audio feature of the first speech and the second audio feature of the second speech are extracted; the audio feature matching degree between the first audio feature and the second audio feature is determined; and the audio feature matching degree is determined as similarity.

[0083] Among them, audio features can include the timbre and pitch of the audio.

[0084] Method 2: Determine the similarity based on the text feature matching degree of the first and second speech.

[0085] Specifically, the first speech is converted into text to obtain the first speech text; the second speech is converted into text to obtain the second speech text; the text feature matching degree between the first speech text and the second speech text is determined; and the text feature matching degree is determined as the similarity.

[0086] In one implementation, automatic speech recognition (ASR) technology can be used for text conversion.

[0087] Method 3: Determine the similarity based on the audio feature matching degree and the text feature matching degree.

[0088] Specifically, the audio feature matching degree of the first speech and the second speech is determined; the text feature matching degree of the first speech and the second speech is determined; and the similarity is determined based on the audio feature matching degree and the text feature matching degree.

[0089] In one implementation, when determining similarity based on audio feature matching degree and text feature matching degree, any of the following methods can be used:

[0090] Method 1: Weighted summation of audio feature matching degree and text feature matching degree to obtain similarity.

[0091] Specifically, the first product between the audio feature matching degree and the first weight is determined, the second product between the text feature matching degree and the second weight is determined, and the sum between the first product and the second product is determined to obtain the similarity.

[0092] In practical applications, the first and second weights can be set according to the actual application scenario, and there are no restrictions here.

[0093] Method 2: The similarity is determined by combining the audio feature matching degree and the text feature matching degree.

[0094] In this way, the similarity between different speech can be determined based on the audio features of the speech and / or the speech text.

[0095] Furthermore, second speech that does not meet the conditions for echo speech can be filtered out. Specifically, any of the following methods can be used:

[0096] Method 1: If the duration of the second speech is determined to be lower than the duration threshold, then the second speech is removed.

[0097] Method 2: Determine the number of words in the second speech text. If the number of words is lower than the word count threshold, then remove the second speech text.

[0098] Since voice text typically consists of simple responses like "um," "okay," or "suffice," the likelihood of it being an echo is low, resulting in low accuracy in echo detection.

[0099] As an example, the calling device belongs to the dialogue management side. It performs text conversion on the second voice to obtain the second voice text. If it is determined that the number of characters in the second voice text is less than five, it is determined that the second voice does not need to be evaluated for echo.

[0100] In practical applications, echo speech conditions can be set according to the actual application scenario, without further restrictions.

[0101] This allows us to remove second speech that is not an echo, reducing the amount of data processing required for echo detection and improving the accuracy of echo detection.

[0102] Step 103: Obtain the echo detection results based on the similarity.

[0103] Specifically, when performing step 103, any of the following methods can be used:

[0104] Method 1: If the similarity is determined to meet the conditions for speech anomalies, then an echo is confirmed to exist.

[0105] Method 2: If the similarity is determined to meet the conditions for speech abnormality, then an echo abnormality is determined to exist, and the number of echo abnormalities is accumulated. If the number of echo abnormalities is determined to be higher than the threshold, then an echo is determined to exist.

[0106] Specifically, the initial value n for the number of echo anomalies is 0. If the similarity is determined to meet the conditions for speech anomalies, n is incremented by one. If the updated n is higher than the number threshold, then an echo is determined to exist.

[0107] As an example, if the similarity threshold is 0.9, the number of echo anomalies is 3, and the number of echo anomalies n is 3, then the first speech and the second speech are collected, and it is determined that the similarity between the first speech and the second speech is 0.95, which is higher than the similarity threshold of 0.9. Then the number of echo anomalies n is updated. If the updated number of echo anomalies n = 4 > 3, then it is determined that there is an echo.

[0108] In one implementation, when determining that the similarity meets the conditions for speech anomalies, any of the following methods can be used:

[0109] Method 1: If the similarity is determined to be higher than the similarity threshold, then an echo is determined to exist.

[0110] Method 2: Similarity includes audio feature matching degree and text feature matching degree. If the audio feature matching degree is determined to be higher than the first matching degree threshold, and / or the text feature matching degree is determined to be higher than the second matching degree threshold, then an echo is determined to exist.

[0111] In practical applications, the similarity threshold, the first matching threshold (e.g., 0.9), and the second matching degree can be set according to the actual application scenario, and no restrictions are imposed here.

[0112] Furthermore, if an echo is detected during a voice call, the echo can be filtered out.

[0113] Specifically, if the echo detection results indicate the presence of an echo, the second speech is filtered.

[0114] In one implementation, the number of echo filtering attempts is determined. If the number of echo filtering attempts exceeds a threshold (e.g., 1), an echo anomaly notification is sent to the first caller. This allows the first caller to explain the echo problem to the second caller when they determine that frequent echoes exist, so that the second caller can change the call location or renegotiate the call time to avoid voice call errors caused by echoes.

[0115] Furthermore, if the second caller determines that an echo exists, they can configure and enable echo cancellation on the target device.

[0116] In voice communication scenarios, if the second caller is in an open room or in an outdoor valley and is making a voice call with the first caller through the target device, the target device may produce an echo. If the first caller is a robot, it is difficult for it to identify, like a human, whether the sound of the communication device is the second caller's voice or an echo. Therefore, in this embodiment, echo detection is performed using voice similarity. When an echo is present, echo filtering and an abnormal echo alarm can be performed, improving the user experience.

[0117] Based on the same inventive concept, this application also provides an echo detection device. Since the principle of the above device and equipment in solving the problem is similar to that of an echo detection method, the implementation of the above device can refer to the implementation of the method, and the repeated parts will not be described again.

[0118] like Figure 2 As shown, it is a structural schematic diagram of an echo detection device provided in an embodiment of this application, including:

[0119] The first acquisition unit 201 is used to acquire the first voice message sent to the target device;

[0120] The second acquisition unit 202 is used to acquire the second voice response from the target device based on the first voice response.

[0121] The determining unit 203 is used to determine the similarity between the first speech and the second speech;

[0122] Unit 204 is used to obtain echo detection results based on similarity.

[0123] In one embodiment, the determining unit 203 is used to:

[0124] Extract the first audio features of the first speech and the second audio features of the second speech;

[0125] Determine the audio feature matching degree between the first audio feature and the second audio feature;

[0126] The degree of audio feature matching is defined as similarity.

[0127] In one embodiment, the determining unit 203 is used to:

[0128] Perform text conversion on the first speech to obtain the first speech text;

[0129] The second speech is converted into text to obtain the second speech text;

[0130] Determine the text feature matching degree between the first and second speech texts;

[0131] The text feature matching degree is defined as similarity.

[0132] In one embodiment, the determining unit 203 is used to:

[0133] Determine the audio feature matching degree between the first and second speech;

[0134] Determine the text feature matching degree between the first and second speech;

[0135] Similarity is determined based on the matching degree of audio features and the matching degree of text features.

[0136] In one embodiment, the determining unit 203 is used to:

[0137] The similarity score is obtained by weighted summation of audio feature matching scores and text feature matching scores.

[0138] Alternatively, the similarity score can be determined by combining the audio feature matching score and the text feature matching score.

[0139] In one embodiment, the obtaining unit 204 is used to:

[0140] If the similarity is determined to meet the conditions for speech anomalies, then an echo is confirmed to exist.

[0141] Alternatively, if the similarity is determined to meet the conditions for speech anomalies, then an echo anomaly is determined to exist, and the number of echo anomalies is accumulated. If the number of echo anomalies is determined to be higher than the threshold, then an echo is determined to exist.

[0142] In one embodiment, the obtaining unit 204 is further configured to:

[0143] If the echo detection results indicate the presence of an echo, then the second speech is filtered.

[0144] The echo detection method, apparatus, electronic device, and computer-readable storage medium provided in this application embodiment acquire a first voice message sent to a target device; acquire a second voice message responded by the target device based on the first voice message; determine the similarity between the first and second voice messages; and obtain an echo detection result based on the similarity. This allows for accurate detection of voice echoes, thus avoiding voice interaction errors caused by echoes.

[0145] Figure 3 A schematic diagram of the structure of an electronic device 3000 is shown. (See also...) Figure 3 As shown, the electronic device 3000 includes a processor 3010 and a memory 3020, and optionally may also include a power supply 3030, a display unit 3040, and an input unit 3050.

[0146] The processor 3010 is the control center of the electronic device 3000. It connects various components through various interfaces and lines, and performs various functions of the electronic device 3000 by running or executing software programs and / or data stored in the memory 3020, thereby performing overall monitoring of the electronic device 3000.

[0147] In this embodiment, when the processor 3010 calls the computer program stored in the memory 3020, it executes the steps in the above embodiments.

[0148] Optionally, the processor 3010 may include one or more processing units; preferably, the processor 3010 may integrate an application processor and a modem processor, wherein the application processor mainly handles the operating system, user interface, and applications, and the modem processor mainly handles wireless communication. It is understood that the modem processor may not be integrated into the processor 3010. In some embodiments, the processor and memory may be implemented on a single chip; in some embodiments, they may also be implemented separately on independent chips.

[0149] The memory 3020 may primarily include a program storage area and a data storage area. The program storage area may store the operating system, various applications, etc.; the data storage area may store data created based on the use of the electronic device 3000, etc. In addition, the memory 3020 may include high-speed random access memory, and may also include non-volatile memory, such as at least one disk storage device, flash memory device, or other volatile solid-state storage device, etc.

[0150] Electronic device 3000 also includes a power supply 3030 (such as a battery) that supplies power to various components. The power supply can be logically connected to processor 3010 through a power management system, thereby enabling the management of charging, discharging, and power consumption.

[0151] The display unit 3040 can be used to display information input by the user or information provided to the user, as well as various menus of the electronic device 3000. In this embodiment of the invention, it is mainly used to display the display interfaces of various applications in the electronic device 3000, as well as text, images, and other objects displayed on the display interfaces. The display unit 3040 may include a display panel 3041. The display panel 3041 may be configured in the form of a liquid crystal display (LCD), an organic light-emitting diode (OLED), or the like.

[0152] The input unit 3050 can be used to receive information such as numbers or characters input by the user. The input unit 3050 may include a touch panel 3051 and other input devices 3052. The touch panel 3051, also known as a touch screen, can collect touch operations on or near the user (such as operations performed by the user using a finger, stylus, or any suitable object or accessory on or near the touch panel 3051).

[0153] Specifically, the touch panel 3051 can detect user touch operations and the signals generated by these operations, convert them into touch point coordinates, send them to the processor 3010, and receive and execute commands from the processor 3010. Furthermore, the touch panel 3051 can be implemented using various types of sensors, including resistive, capacitive, infrared, and surface acoustic wave sensors. Other input devices 3052 can include, but are not limited to, one or more of the following: physical keyboard, function keys (such as volume control buttons, power buttons, etc.), trackball, mouse, joystick, etc.

[0154] Of course, the touch panel 3051 can cover the display panel 3041. When the touch panel 3051 detects a touch operation on or near it, it transmits the information to the processor 3010 to determine the type of touch event. Subsequently, the processor 3010 provides corresponding visual output on the display panel 3041 according to the type of touch event. Although in Figure 3 In this embodiment, the touch panel 3051 and the display panel 3041 are two separate components to realize the input and output functions of the electronic device 3000. However, in some embodiments, the touch panel 3051 and the display panel 3041 can be integrated to realize the input and output functions of the electronic device 3000.

[0155] The electronic device 3000 may also include one or more sensors, such as a pressure sensor, a gravity acceleration sensor, a proximity sensor, etc. Of course, depending on the specific application, the electronic device 3000 may also include other components such as a camera. Since these components are not the focus of this application's embodiments, therefore... Figure 3It is not shown in the text and will not be described in detail here.

[0156] Those skilled in the art will understand that Figure 3 This is merely an example of an electronic device and does not constitute a limitation on the electronic device. It may include more or fewer components than shown, or a combination of certain components, or different components.

[0157] In this embodiment of the application, a computer-readable storage medium stores a computer program thereon. When the computer program is executed by a processor, it enables a communication device to perform the various steps in the above embodiments.

[0158] For ease of description, the above sections are divided into modules (or units) according to their functions and described separately. Of course, in implementing this application, the functions of each module (or unit) can be implemented in one or more software or hardware components.

[0159] Those skilled in the art will understand that embodiments of this application can be provided as methods, systems, or computer program products. Therefore, this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this application can take the form of a computer program product embodied on one or more computer-usable storage media (including but not limited to disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code.

[0160] This application is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of this application. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart... Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0161] These computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to function in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The function specified in one or more boxes.

[0162] These computer program instructions may also be loaded onto a computer or other programmable data processing equipment to cause a series of operational steps to be performed on the computer or other programmable equipment to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable equipment for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0163] Although preferred embodiments of this application have been described, those skilled in the art, upon learning the basic inventive concept, can make other changes and modifications to these embodiments. Therefore, the appended claims are intended to be interpreted as including the preferred embodiments as well as all changes and modifications falling within the scope of this application.

[0164] Obviously, those skilled in the art can make various modifications and variations to this application without departing from the spirit and scope of this application. Therefore, if such modifications and variations fall within the scope of the claims of this application and their equivalents, this application also intends to include such modifications and variations.

Claims

1. A method for echo detection, characterized in that, include: Acquire the first voice message sent to the target device; Obtain the second voice response from the target device based on the first voice response; Filter out second speech that does not meet the conditions for echo speech; specifically, if the duration of the second speech is determined to be lower than the duration threshold, then remove the second speech; or, determine the number of words in the second speech text, and if the number of words is lower than the number of words threshold, then remove the second speech. Determine the similarity between the first speech and the second speech; Based on the similarity, the echo detection result is obtained.

2. The method as described in claim 1, characterized in that, Determining the similarity between the first speech and the second speech includes: Extract the first audio features of the first speech and the second audio features of the second speech; Determine the audio feature matching degree between the first audio feature and the second audio feature; The audio feature matching degree is determined as the similarity.

3. The method as described in claim 1, characterized in that, Determining the similarity between the first speech and the second speech includes: The first speech is converted into text to obtain the first speech text; The second speech is converted into text to obtain the second speech text. Determine the text feature matching degree between the first speech text and the second speech text; The text feature matching degree is determined as the similarity.

4. The method according to any one of claims 1-3, characterized in that, Determining the similarity between the first speech and the second speech includes: Determine the audio feature matching degree between the first speech and the second speech; Determine the text feature matching degree between the first speech and the second speech; The similarity is determined based on the audio feature matching degree and the text feature matching degree.

5. The method as described in claim 4, characterized in that, Determining the similarity based on the audio feature matching degree and the text feature matching degree includes: The similarity is obtained by weighted summing of the audio feature matching score and the text feature matching score; Alternatively, the similarity can be determined by combining the audio feature matching degree and the text feature matching degree.

6. The method as described in claim 5, characterized in that, The step of obtaining the echo detection result based on the similarity includes: If the similarity is determined to meet the conditions for speech anomalies, then an echo is determined to exist; Alternatively, if the similarity is determined to meet the speech anomaly conditions, then an echo anomaly is determined to exist, and the number of echo anomalies is accumulated. If the number of echo anomalies is determined to be higher than the number threshold, then an echo is determined to exist.

7. The method according to any one of claims 1-3, characterized in that, The method further includes: If an echo is determined to exist based on the echo detection results, the second speech is then filtered.

8. An echo detection device, characterized in that, include: The first acquisition unit is used to acquire the first voice message sent to the target device; The second acquisition unit is used to acquire the second voice response from the target device based on the first voice response; The second acquisition unit is also used to filter out second speech that does not meet the conditions for echo speech; specifically, it includes: if the duration of the second speech is determined to be lower than the duration threshold, then the second speech is removed; or, if the number of characters in the second speech text is determined, and the number of characters is lower than the number of characters threshold, then the second speech is removed. A determining unit is configured to determine the similarity between the first speech and the second speech; The obtaining unit is used to obtain the echo detection result based on the similarity.

9. The apparatus as claimed in claim 8, characterized in that, The determining unit is used for: Extract the first audio features of the first speech and the second audio features of the second speech; Determine the audio feature matching degree between the first audio feature and the second audio feature; The audio feature matching degree is determined as the similarity.

10. The apparatus as claimed in claim 8, characterized in that, The determining unit is used for: The first speech is converted into text to obtain the first speech text; The second speech is converted into text to obtain the second speech text. Determine the text feature matching degree between the first speech text and the second speech text; The text feature matching degree is determined as the similarity.

11. The apparatus according to any one of claims 8-10, characterized in that, The determining unit is used for: Determine the audio feature matching degree between the first speech and the second speech; Determine the text feature matching degree between the first speech and the second speech; The similarity is determined based on the audio feature matching degree and the text feature matching degree.

12. The apparatus as claimed in claim 11, characterized in that, The determining unit is used for: The similarity is obtained by weighted summing of the audio feature matching score and the text feature matching score; Alternatively, the similarity can be determined by combining the audio feature matching degree and the text feature matching degree.

13. The apparatus as claimed in claim 12, characterized in that, The obtaining unit is used for: If the similarity is determined to meet the conditions for speech anomalies, then an echo is determined to exist; Alternatively, if the similarity is determined to meet the speech anomaly conditions, then an echo anomaly is determined to exist, and the number of echo anomalies is accumulated. If the number of echo anomalies is determined to be higher than the number threshold, then an echo is determined to exist.

14. The apparatus as claimed in claim 13, characterized in that, The obtaining unit is further configured to: If an echo is determined to exist based on the echo detection results, the second speech is then filtered.

15. An electronic device, characterized in that, It includes a processor and a memory, the memory storing computer-readable instructions that, when executed by the processor, perform the method as described in any one of claims 1-7.

16. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by a processor, it performs the method as described in any one of claims 1-7.

Images