Articulation detection method and apparatus, electronic device, storage medium

Abstract

The embodiment of the present disclosure provides a clipping detection method and device, electronic equipment and storage medium, and belongs to the technical field of signal processing. The method comprises the following steps: dividing a to-be-detected audio signal into N time-domain signal frames; determining a first clipping confidence degree corresponding to each time-domain signal frame according to the statistical distribution of signal amplitudes in the frame, wherein the first clipping confidence degree is used to distinguish normal audio signals and clipping audio signals; determining a second clipping confidence degree corresponding to a detection window formed by M continuous time-domain signal frames according to the first clipping confidence degrees of the M continuous time-domain signal frames; and determining that there is clipping in the corresponding detection window in response to the second clipping confidence degree being greater than a preset first threshold. The present disclosure solves the problems of high false detection rate and low accuracy of clipping detection caused by the existing fixed clipping detection threshold.

Description

TECHNICAL FIELD

[0001] The present disclosure relates to the technical field of signal processing, and more particularly, embodiments of the present disclosure relate to a popping detection method and apparatus, electronic device, and storage medium. BACKGROUND

[0002] This section is intended to provide background or context to the embodiments of the disclosure recited in the claims. The description herein does not constitute admission that the background art is prior art to the present disclosure.

[0003] Popping, also known as popping, is a very harsh sound, and in the presence of popping, the original sound will become unclear. When there is popping in the audio signal, the audio quality will be severely affected, therefore, it is necessary to detect the popping of the audio signal.

[0004] Since most of the popping is caused by signal clipping, in the related art, a clipping threshold is usually used to detect clipping, and then clipping is used to determine popping. However, in fact, clipping does not necessarily produce audible popping, which can easily lead to false detection; and for a signal with large fluctuations, the accuracy of detection using a fixed clipping threshold is low. SUMMARY

[0005] Embodiments of the present disclosure provide a popping detection method and apparatus, electronic device, and storage medium.

[0006] In a first aspect of the embodiments of the present disclosure, a popping detection method is provided, the method comprising: dividing a to-be-detected audio signal into N time-domain signal frames, the N being a positive integer; determining a first clipping confidence corresponding to each time-domain signal frame according to a statistical distribution of signal amplitudes in the frame, the first clipping confidence being used to distinguish between normal audio signals and clipped audio signals; determining a second clipping confidence corresponding to a detection window formed by M consecutive time-domain signal frames according to the first clipping confidence of the M consecutive time-domain signal frames, the M being a positive integer smaller than the N; and in response to the second clipping confidence being greater than a preset first threshold, determining that there is popping in the corresponding detection window.

[0007] Optionally, the determining the first clipping confidence according to the statistical distribution of the amplitude values of each frame of the time-domain signal comprises: for each frame of the time-domain signal, counting the number of signals in each amplitude section to construct a statistical histogram, the amplitude section being obtained by dividing the amplitude interval formed by the maximum amplitude and the minimum amplitude of the frame of the time-domain signal; searching for a target block in the statistical histogram, the target block being a block in which the two ends are higher than the middle and the back end is higher than the front end, the front end and the back end being determined according to the searching order; determining the horizontal distance of the two ends of each target block; and determining the first clipping confidence according to the ratio of the maximum value in the horizontal distances to the total number of the amplitude sections.

[0008] Optionally, the searching for the target block comprises: starting from the two ends of the statistical histogram respectively, moving to the middle in turn, and searching for the target block.

[0009] Optionally, the method further comprises: for each frame of the time-domain signal, determining a first clipping ratio of the frame according to the proportion of a target signal in the frame of the time-domain signal, the target signal being a signal corresponding to the maximum amplitude value and the minimum amplitude value; determining a second clipping ratio of a corresponding detection window according to the first clipping ratios of M consecutive frames of the time-domain signal; and the determining that the detection window contains the broken tone comprises: in response to the second clipping confidence being greater than a preset first threshold and the second clipping ratio being greater than a preset second threshold, determining that the detection window contains the broken tone.

[0010] Optionally, the method further comprises: determining a frequency domain energy feature of the frames of the time-domain signal in the detection window; and the determining that the detection window contains the broken tone comprises: in response to the frames of the signal in the detection window not satisfying at least one of the following conditions: the second clipping confidence being greater than the preset first threshold and the second clipping ratio being greater than the preset second threshold, determining whether the detection window contains the broken tone according to the frequency domain energy feature.

[0011] Optionally, the frequency domain energy feature comprises a cutoff frequency value, and the determining the frequency domain energy feature of the frames of the time-domain signal in the detection window comprises: performing time-frequency transformation on each frame of the time-domain signal in the detection window to obtain a corresponding frequency domain signal; and determining the energy center of the frequency domain signal as the cutoff frequency value of the signal frame.

[0012] Optionally, the determining whether the detection window contains the broken tone according to the frequency domain energy feature comprises: in response to the cutoff frequency value of the signal frame being greater than a frequency threshold value, determining that the signal frame contains clipping, the frequency threshold value being determined based on the maximum frequency value of the frequency domain signal; and in response to there being at least a preset number of signal frames with clipping in the detection window, determining that the detection window contains the broken tone.

[0013] Optionally, the determining the second clipping confidence corresponding to the M time-domain signal frames according to the first clipping confidence of the M time-domain signal frames comprises: performing first weighting processing on the first clipping confidence of the M time-domain signal frames in the detection window to obtain the second clipping confidence; and the determining the second clipping ratio of the detection window according to the first clipping ratio of the M time-domain signal frames comprises: performing second weighting processing on the first clipping ratio of the M time-domain signal frames in the detection window to obtain the second clipping ratio.

[0014] Optionally, the M is determined based on a minimum time length of clipping forming a glitch and a length of a time-domain signal frame.

[0015] In a second aspect of the embodiments of the present application, a glitch detection device is provided, characterized in that the device comprises: a signal division module configured to divide an audio signal to be detected into N time-domain signal frames, the N being a positive integer; a first determination module configured to determine a first clipping confidence corresponding to each time-domain signal frame according to a statistical distribution of signal amplitudes in the frame, the first clipping confidence being used to distinguish normal audio signals and clipping audio signals; a second determination module configured to determine a second clipping confidence corresponding to a detection window formed by M time-domain signal frames according to the first clipping confidence of the M time-domain signal frames, the M being a positive integer smaller than the N; and a glitch determination module configured to determine that there is a glitch in the detection window corresponding in response to the second clipping confidence being greater than a preset first threshold.

[0016] Optionally, the first determination module comprises: a histogram construction module configured to, for each time-domain signal frame, count the number of signals in each amplitude section to construct a statistical histogram, the amplitude section being obtained by dividing an amplitude interval formed by a maximum amplitude and a minimum amplitude of the time-domain signal frame; a searching module configured to search for a target block in the statistical histogram, the target block being a block in which both ends of the block are higher than the middle of the block, and the back end of the block is higher than the front end of the block, the front end of the block and the back end of the block being determined according to the order of searching; a distance determination module configured to determine a horizontal distance of both ends of each target block; and a confidence determination module configured to determine the first clipping confidence according to a ratio of a maximum value in the horizontal distances to a total number of the amplitude sections.

[0017] Optionally, the searching module is further configured to: start from both ends of the statistical histogram respectively, and sequentially move to the middle to search for the target block.

[0018] Optionally, the apparatus further comprises a clipping ratio determination module configured to: for each frame of the time-domain signal frames, determine a first clipping ratio of the frame according to a proportion of a target signal in the frame, the target signal being a signal corresponding to the maximum amplitude value and the minimum amplitude value; determine a second clipping ratio of a detection window according to the first clipping ratios of the M continuous time-domain signal frames; and the clipping determination module is further configured to: in response to the second clipping confidence being greater than a preset first threshold and the second clipping ratio being greater than a preset second threshold, determine that there is clipping in the detection window.

[0019] Optionally, the apparatus further comprises a frequency domain feature determination module configured to determine a frequency domain energy feature of the time-domain signal frames in the detection window; and the clipping determination module is further configured to: in response to the signal frames in the detection window not satisfying at least one of the following: the second clipping confidence being greater than the preset first threshold and the second clipping ratio being greater than the preset second threshold, determine whether there is clipping in the detection window according to the frequency domain energy feature.

[0020] Optionally, the frequency domain energy feature comprises a cutoff frequency value, and the frequency domain feature determination module is further configured to: perform time-frequency transformation on each frame of the time-domain signal frames in the detection window to obtain a corresponding frequency domain signal; and determine an energy center of the frequency domain signal as the cutoff frequency value of the signal frame.

[0021] Optionally, the clipping determination module is further configured to: in response to the cutoff frequency value of the signal frame being greater than a frequency threshold value, determine that the signal frame has clipping, the frequency threshold value being determined based on a maximum frequency value of the frequency domain signal; and in response to there being at least a preset number of signal frames with clipping in the detection window, determine that the detection window has clipping.

[0022] Optionally, the second determination module is further configured to: perform first weighting processing on the first clipping confidence of the M time-domain signal frames in the detection window to obtain the second clipping confidence; and the clipping ratio determination module is further configured to: perform second weighting processing on the first clipping ratio of the M time-domain signal frames in the detection window to obtain the second clipping ratio.

[0023] Optionally, the M is determined based on a minimum time length of continuous clipping forming clipping and a length of the time-domain signal frame.

[0024] In a third aspect of the embodiments of the present application, a storage medium is provided, and the storage medium has a program stored thereon, the program being executed by a processor to implement the method in the above embodiments.

[0025] In a fourth aspect of the embodiments of the present application, an electronic device is provided, comprising a processor and a memory, the memory storing executable instructions, and the processor is configured to invoke the executable instructions stored in the memory to execute the method in the above embodiments.

[0026] According to the clipping detection method provided by the embodiments of the present application, on the one hand, the first clipping confidence of each frame of time domain signal can be determined according to the statistical distribution of the signal amplitude in the frame, so as to distinguish the normal audio signal and the clipping audio signal, that is, the clipping threshold can be dynamically adjusted based on the statistical distribution of the signal, thereby solving the problem of the clipping detection accuracy caused by the fixed clipping detection threshold, and further ensuring the accuracy of the clipping detection. On the other hand, the second clipping confidence of the detection window is determined by the first clipping confidence of the continuous M frames of time domain signals, so as to detect the clipping with the detection window as the detection unit, so that the detected clipping is actually the perceptual clipping, the false detection phenomenon of the short-time clipping without perceptual clipping is avoided, and the clipping false detection rate is reduced. In addition, the clipping detection is performed in the time domain, and the calculation complexity is low and the detection efficiency is high. BRIEF DESCRIPTION OF DRAWINGS

[0027] The above and other objects, features and advantages of the exemplary embodiments of the present application will be more apparent from the following detailed description taken in conjunction with the accompanying drawings, in which:

[0028] FIG. 1 A flowchart of a clipping detection method according to one embodiment of the present application is schematically shown.

[0029] FIG. 2 A schematic diagram of a statistical histogram according to one embodiment of the present application is schematically shown.

[0030] FIG. 3 A flowchart of determining the first clipping confidence of a signal frame based on a statistical histogram according to one embodiment of the present application is schematically shown.

[0031] FIG. 4 A flowchart of a clipping detection process according to one embodiment of the present application is schematically shown.

[0032] FIG. 5 A clipping detection result diagram of an audio signal according to one embodiment of the present application is schematically shown.

[0033] FIG. 6 A block diagram of a clipping detection device structure according to one embodiment of the present application is schematically shown.

[0034] FIG. 7A structural diagram of an electronic device suitable for implementing embodiments of the present application is schematically shown.

[0035] In the drawings, identical or corresponding numbers represent identical or corresponding parts. DETAILED DESCRIPTION

[0036] The principles and spirits of the present disclosure will be described below with reference to several exemplary embodiments. It should be understood that these embodiments are given only so that those skilled in the art can better understand and implement the present disclosure, and do not limit the scope of the present disclosure in any way. On the contrary, these embodiments are provided so that the present disclosure is more thorough and complete, and the scope of the present disclosure is fully conveyed to those skilled in the art.

[0037] Those skilled in the art know that the embodiments of the present disclosure can be implemented as a system, a device, an apparatus, a method or a computer program product. Therefore, the present disclosure can be embodied as a complete hardware, a complete software (including firmware, resident software, microcode, etc.), or a combination of hardware and software.

[0038] According to embodiments of the present disclosure, a clip detection method and device, an electronic device, and a storage medium are provided. SUMMARY

[0040] Clip: also known as pop, is a phenomenon that the input audio signal exceeds the maximum range of the current digital representation of the audio signal, resulting in a significant impairment of the sound quality, which is manifested as a harsh sound, making the original sound unclear. From the spectrogram, the energy of each frequency band is very high.

[0041] Clipping: also known as amplitude limiting, is a form of audio distortion, which is limited when the signal exceeds a certain threshold. From the waveform, the peaks and troughs are flattened, and the numerical value becomes the threshold. Generally, the threshold of recording equipment is set to 1 to prevent excessive power. However, clipping to a certain extent will lead to clip in the sense of hearing.

[0042] Fast Fourier Transform (FFT): is a fast algorithm of discrete Fourier transform, which is used to transform the time domain signal to the frequency domain. IFFT is the inverse transform of FFT, i.e. transforming the frequency domain signal back to the time domain.

[0043] Mel Frequency (Mel spectrum): refers to a two-dimensional signal form (spectrum) obtained by dividing a long signal into frames, windowing, and then performing FFT on each frame based on short-time Fourier transform.

[0044] The cause of the clipped sound can be that the volume of the audio signal exceeds the upper limit of the analog-to-digital converter of the recording device during the process of collecting the audio by the microphone. The cause of the clipped sound can also be that the audio signal is amplified during the process of processing the audio signal, and the amplification result exceeds the maximum digital representation of the sound bit depth. In terms of listening, the clipped sound is manifested as hoarse or noisy sound.

[0045] Clipped sound detection is to detect whether there is a clipped sound phenomenon in the audio signal and the specific position of the clipped sound. Through the clipped sound detection, the recording parameters or the audio signal processing parameters can be adjusted in time to avoid the occurrence of the clipped sound. The clipped sound detection can also be used to assist the clipped sound repair algorithm to repair the detected position of the clipped sound.

[0046] In the related art, according to the domain of analysis, the clipped sound detection can be divided into time domain clipped sound detection and frequency domain clipped sound detection. In the time domain, the values of each sampling point of the audio are mainly counted. In the frequency domain, the features are extracted by calculating the spectrum to analyze and judge. However, whether the time domain method or the frequency domain method, the clipped sound is detected by detecting whether there is a clipping segment and the position of the clipping segment. In fact, short-time clipping does not cause listening clipped sound, that is, the listening clipped sound is not completely equivalent to clipping.

[0047] To solve the above problems, the inventive concept of the present application is to determine the first clipping confidence of a frame by the signal statistical distribution in the time domain signal frame, to perform the clipping detection of the frame, so as to realize the dynamic clipping detection threshold of different frames, and to improve the clipping detection accuracy. A clipped sound detection window formed by M continuous time domain signal frames is also designed. The clipped sound detection is performed in units of the detection window, so as to reduce the false detection rate of short-time clipping as a clipped sound. The frame features and the continuous frame features of the continuous multiple frames are combined to realize the accurate detection of the listening clipped sound.

[0048] The clipped sound detection method of the present application can be used in the audio processing process of various application programs. For example, the application programs with audio processing functions such as social application programs, music application programs, video application programs, and game application programs. For the social application program, the clipped sound detection method provided by the present application can be used to detect whether the voice in the social communication process has a clipped sound and the position of the clipped sound in real time, and to adjust the device parameters or the network state in time. For the music application program, the clipped sound detection method provided by the present application can be used to detect whether the played music has a clipped sound and the position of the clipped sound, to adjust the device parameters or switch other audio, and to ensure the user experience.

[0049] The method and device of the embodiments of the present disclosure can be applied to at least one of electronic devices including but not limited to a server, a terminal, and the like, which can be configured to execute the method provided by the embodiments of the present disclosure. In other words, the clipping detection method can be executed by software or hardware installed in a terminal device or a server device, and the software can be a blockchain platform. The server includes but is not limited to a single server, a server cluster, a cloud server, or a cloud server cluster, and the like. The present disclosure takes an electronic device as an example for illustration.

[0050] Exemplary methods

[0051] The preferred embodiments of the present disclosure are described below in conjunction with the accompanying drawings. It should be understood that the preferred embodiments described herein are only used to illustrate and explain the present disclosure, and are not used to limit the present disclosure, and the embodiments in the present disclosure and the features in the embodiments can be combined with each other without conflict.

[0052] The clipping detection method according to the exemplary embodiments of the present disclosure will be described below with reference to FIG. 1 may include steps S110-S140.

[0053] In step S110, the audio signal to be detected is divided into N time domain signal frames, and N is a positive integer.

[0054] In step S120, according to the statistical distribution of the signal amplitude in each frame of the time domain signal frame, the first clipping confidence corresponding to the frame is determined, and the first clipping confidence is used to distinguish between normal audio signals and clipping audio signals.

[0055] In step S130, according to the first clipping confidence of the continuous M time domain signal frames, the second clipping confidence corresponding to the detection window formed by the continuous M time domain signal frames is determined, and M is a positive integer less than N.

[0056] In step S140, in response to the second clipping confidence being greater than a preset first threshold, it is determined that there is a clipping in the corresponding detection window.

[0057] The method for detecting clipped sound provided by the embodiments of the present disclosure can determine the first clipping confidence corresponding to each time-domain signal frame according to the statistical distribution of the signal amplitude in the frame, so as to distinguish normal audio signals and clipped audio signals, that is, the clipping threshold can be dynamically adjusted based on the statistical distribution of the signal, thereby solving the problem of the clipping detection accuracy caused by the fixed clipping detection threshold, and further ensuring the accuracy of the clipped sound detection. On the other hand, the second clipping confidence of the detection window is determined by the first clipping confidence of the continuous M time-domain signal frames, and the clipped sound is detected by taking the detection window as the detection unit, so that the detected clipped sound is actually the audible clipped sound, the false detection phenomenon of short-time clipping without audible clipped sound is avoided, and the clipped sound false detection rate is reduced. In addition, the present application detects the clipped sound in the time domain, has low computational complexity and high detection efficiency.

[0058] The above steps are described in detail below.

[0059] In step S110, the audio signal to be detected is divided into N time-domain signal frames.

[0060] In the present example embodiment, the audio signal to be detected can be a time-domain audio signal, which can be a recorded audio, a real-time call audio, or a music audio to be played, etc., which is not limited in the present example. The audio signal to be detected can be divided into frames according to a fixed time length, such as 5 milliseconds, 10 milliseconds, 15 milliseconds, etc., or the signal frames can be divided according to a set frame length and overlap rate, which is not limited in the present example. The overlap rate refers to the degree of signal overlap, and the signal overlap refers to the existence of a certain length of overlap between adjacent two signal frames. The signal overlap can eliminate a part of the real signal consumed by the window function weighting. N is the number of divided signal frames, which can be an integer of 1 or greater. The time stamp of each signal frame can also be recorded to distinguish the signal frames and their order.

[0061] In step S120, the first clipping confidence corresponding to each time-domain signal frame is determined according to the statistical distribution of the signal amplitude in the frame.

[0062] In the present example embodiment, for each time-domain signal frame, the distribution of the signal amplitude can be counted, and the first clipping confidence is determined based on the counting result. The first clipping confidence is used to distinguish normal audio signals and clipped audio signals.

[0063] Exemplarily, the first clipping confidence can be determined by the following steps.

[0064] Firstly, for each time-domain signal frame, the number of signals in each amplitude section is counted to construct a statistical histogram.

[0065] In this example implementation, the amplitude interval formed by the maximum and minimum amplitudes of the time-domain signal frame can be segmented. For example, the amplitude interval can be divided into K equal segments (K is a positive integer), the number of signals in each amplitude segment can be counted, and a statistical histogram can be constructed, such as... FIG. 2 As shown, the horizontal axis of the histogram represents the amplitude segment number (e.g., 1, 2, ... K), and the vertical axis represents the number of signals.

[0066] The second step is to find the target block in the statistical histogram.

[0067] In this example implementation, the target block refers to a block in the statistical histogram where both ends are higher than the middle, and the rear end is higher than the front end. The front and rear ends are determined by the order in which the statistical histogram is traversed; the histogram bars traversed first are the front ends, and the histogram bars traversed later are the rear ends. The target block is as follows: FIG. 2 The dashed box in the diagram illustrates this. The target block can be found by iterating through the histogram. For example, starting from the left end of the histogram and moving towards the right, the target block can be found where both ends are higher than the middle, and the right end is higher than the left. Alternatively, the target block can be found by moving from the right end of the histogram and moving towards the left. For instance, the search can also be accelerated by starting from both ends (left and right) and moving sequentially towards the middle.

[0068] The third step is to determine the lateral distance between the two ends of each target block.

[0069] In this example implementation, the horizontal distance can be obtained by subtracting the number at the right end of the block from the number at the left end of the block, such as... FIG. 2 The horizontal distance in the middle is k2-k1, 1≤k1 <k2≤K。

[0070] The fourth step is to determine the first clipping confidence level based on the ratio of the maximum value in each lateral distance to the total number of amplitude segments.

[0071] In this example implementation, when the number of target blocks is 1, the lateral distance of the target block can be divided by the amplitude range to determine the first clipping confidence level. When the number of target blocks is greater than 1, the target block with the largest lateral distance can be selected to determine the first clipping confidence level.

[0072] For example, suppose that for the k-th amplitude segment of K amplitude segments, the number of signals falling within that segment is H(k). FIG. 3 As shown, the first clipping confidence level can be determined through the following steps.

[0073] Step S301, initialize the left histogram number Yl0=H(kl), the right histogram number Yr0=H(kr), the left distance dl (the horizontal distance of the left pointer from the left end of the histogram) and the right distance dr (the horizontal distance of the right pointer from the right end of the histogram) as 0, i.e. dl=dr=0, the left pointer kl=1, the right pointer kr=K; the maximum horizontal distance parameter Dmax=0.

[0074] Step S302, increase the left pointer kl by 1 and the left distance dl by 1.

[0075] Step S303, determine whether the left histogram number H(kl) corresponding to the kl pointer in the statistical histogram is greater than Yl0, if yes, go to S304, otherwise go to S308.

[0076] Step S304, update Yl0=H(kl) and reset dl as 0, otherwise, go to step S340. This step is the process of traversing from the left to the middle.

[0077] Step S305, decrease the right pointer kr by 1 and increase the right distance dr by 1.

[0078] Step S306, determine whether the right histogram number H(kr) corresponding to the kr pointer in the statistical histogram is greater than Yr0, if yes, go to S307, otherwise go to S308.

[0079] Step S307, update Yr0=H(kr) and reset dr as 0, otherwise, go to step S340. This step is the process of traversing from the right to the middle.

[0080] Step S308, update Dmax as the maximum value among Dmax, dl and dr at this time.

[0081] Step S309, determine whether kr is less than kl, if yes, go to step S310, otherwise go to step S302 and S305 to enter the next loop.

[0082] Step S310, determine the first clipping confidence Rcl of the time frame as Dmax / K.

[0083] In the above steps, S302-S303 and S305-306 can be performed synchronously, and the first clipping confidence of each signal frame can be determined and recorded according to the above process.

[0084] In step S130, the second clipping confidence corresponding to the detection window formed by the continuous M time domain signal frames is determined according to the first clipping confidence of the continuous M time domain signal frames.

[0085] In the example embodiment, M is a positive integer less than N, and M can be determined based on a minimum length of a time domain signal frame in which a clip artifact is formed by consecutive clipping and a length of the time domain signal frame. The minimum length of the time domain signal frame in which the clip artifact is formed by the consecutive clipping can be determined according to experience and actual conditions. For example, the minimum length can be set to 0.4 seconds. The ratio between the minimum length and the length of the signal frame can be rounded up to obtain M. The ratio can also be appropriately amplified and rounded up to obtain M. The example does not limit the ratio.

[0086] In the example embodiment, the first clip confidence of the consecutive M time domain signal frames can be subjected to a correlation operation to determine a second clip confidence. For example, the correlation operation can be an average operation, a weighted operation, or an aggregation process. The example does not limit the correlation operation.

[0087] For example, the first clip confidence of the M time domain signal frames in the detection window can be subjected to a first weighted process to obtain the second clip confidence.

[0088] In the example embodiment, the first weighted process is a weighted summation process. The sum of the weights corresponding to the M signal frames can be 1. For example, the first clip confidence of the M signal frames can be subjected to a Gaussian weighted process to obtain the second clip confidence. The Gaussian weighted process means that the values at different positions are calculated with different weights when the average is calculated. The Gaussian weight represents that the weight values follow a Gaussian distribution, with small values at both ends and a large value in the middle.

[0089] In step S140, in response to the second clip confidence being greater than a preset first threshold, it is determined that there is a clip artifact in the corresponding detection window.

[0090] In the example embodiment, the first threshold is a clip artifact threshold of the detection window, which can be set according to actual conditions. In the case where the second clip confidence of the detection window is greater than the first threshold, it can be considered that the detection window has a clip artifact.

[0091] In some embodiments, in addition to the clip confidence, a clip ratio can also be added as a clip artifact detection feature. The method can further include the following steps.

[0092] For each time domain signal frame, a first clip ratio of the frame is determined according to a proportion of the target signal in the frame. A second clip ratio of the corresponding detection window is determined according to the first clip ratio of the consecutive M time domain signal frames.

[0093] In the example embodiment, the target signal is a signal corresponding to the maximum amplitude value and the minimum amplitude value. Since the essence of clipping is that the signal absolute value higher than the threshold value is clipped to the threshold value, the clipping segment must have multiple sampling points with the value equal to the threshold value, and the first clipping ratio is determined based on this. For each time domain signal frame, the maximum amplitude value and the minimum amplitude value are first determined, the number of sampling points n1 of the maximum amplitude value and the minimum amplitude value in the frame is counted, and the first clipping ratio of the frame is n1 / n2, where n2 is the total number of sampling points of the frame.

[0094] In the example embodiment, the first clipping ratios of the continuous M time domain signal frames can be correlated to determine the second clipping ratio. For example, the correlation operation can be an average value operation, a weighted operation or an aggregation process, which is not limited in the example. For example, the second weighting process can be performed on the first clipping ratios of the M time domain signal frames in the detection window to obtain the second clipping ratio. The second weighting process can be a Gaussian weighting process, and the first and second weighting processes can be the same or different, which is not limited in the example.

[0095] The clip confidence and the clip ratio are two features based on which the break sound detection strategy is as follows: in response to the second clip confidence being greater than a preset first threshold value and the second clip ratio being greater than a preset second threshold value, it is determined that there is a break sound in the detection window.

[0096] In the example embodiment, the second threshold value is a threshold value of the break sound in the clip ratio dimension, which can be set according to actual conditions. By determining whether there is a break sound in the detection window based on the two features of the clip confidence and the clip ratio, the detection accuracy of the break sound can be improved.

[0097] In practice, in addition to the obvious clipping segment, there are some scenarios in which the signal value is close to the clipping threshold value but the actual listening experience has a break sound. For this scenario, the present application adds a frequency domain feature for detection based on the characteristics of clipping, that is, clipping will cause higher high-frequency energy than normal audio.

[0098] In some embodiments, the method further comprises: determining a frequency domain energy feature of the time domain signal frame in the detection window.

[0099] In the example embodiment, the frequency domain energy feature can include various energy indicators in the frequency domain, such as power value, energy maximum value, energy center of gravity, cutoff frequency value, etc., which are not limited in the example.

[0100] For example, the frequency domain energy feature includes the cutoff frequency value, the time-frequency transformation can be performed on each time domain signal frame in the detection window to obtain the corresponding frequency domain signal; and then the energy center of gravity of the frequency domain signal is determined as the cutoff frequency value of the signal frame.

[0101] In the example embodiment, the windowing processing and the FFT can be sequentially performed on each signal frame to obtain a corresponding Mel spectrum, and then the energy distribution center, i.e., the frequency Fc, of the Mel spectrum is determined. The energy distribution center can be determined by the ratio of the sum of the energy of each frequency position (the product of the amplitude and the frequency value of the position) to the total number of frequency positions.

[0102] In a case where the signal frames in a detection window do not satisfy: the second clipping confidence is greater than the preset first threshold or / and the second clipping ratio is greater than the preset second threshold, whether there is a glissando in the detection window can be determined based on the cutoff frequency value.

[0103] Exemplarily, the cutoff frequency value of the signal frame can be compared with a frequency threshold first, and in a case where the cutoff frequency value is greater than the frequency threshold, it is determined that the signal frame has clipping. The frequency threshold can be determined based on the maximum frequency value of the frequency domain signal, for example, the frequency threshold can be set to 1 / 3-1 / 4 of the maximum frequency value. Then, whether the signal frames with clipping in the detection window reach a preset number (such as 2) is determined, and in a case where the preset number is reached, it is determined that the detection window has a glissando. Through the frequency energy feature, the special case of no clipping but audible glissando is avoided, and the detection accuracy is improved.

[0104] As shown in FIG. 1, the specific process of a glissando detection method of an embodiment of the present application is introduced as follows. FIG. 4

[0105] Step S401, the audio signal to be detected is divided into N time domain signal frames.

[0106] Step S402, the first clipping confidence corresponding to each time domain signal frame is determined according to the statistical distribution of the signal amplitude in the frame.

[0107] Step S403, the first clipping confidence of the M time domain signal frames in the detection window is subjected to first weighting processing to obtain the second clipping confidence.

[0108] Step S404, for each time domain signal frame, the first clipping ratio of the frame is determined according to the proportion of the target signal in the frame, and the target signal is the signal corresponding to the maximum amplitude value and the minimum amplitude value.

[0109] Step S405, the first clipping ratio of the M time domain signal frames in the detection window is subjected to second weighting processing to obtain the second clipping ratio.

[0110] Step S406, time-frequency transformation is performed on each time domain signal frame in the detection window to obtain a corresponding frequency domain signal.

[0111] Step S407, the energy center of the frequency domain signal is determined as the cutoff frequency value of the signal frame. ​

[0112] Step S408, it is judged whether the second clipping confidence of the current detection window is greater than the first threshold and the second clipping ratio is greater than the second threshold, if yes, go to step S409, otherwise go to step S410.

[0113] Step S409, it is determined that there is a break in the current detection window.

[0114] Step S410, it is judged whether the number of target signal frames in the current detection window is greater than the preset number, if yes, go to S409, otherwise, go to step S411. The target signal frame is a signal frame with a cutoff frequency value greater than a frequency threshold.

[0115] Step S411, slide one step (such as one time frame) forward, go to step S402, and perform break detection on the next detection window.

[0116] In the above embodiment, the determination processes of the break judgment features: clipping confidence (S402, S403), clipping ratio (S404, S405) and frequency domain energy feature (S406, S407) can be performed simultaneously to speed up the detection efficiency.

[0117] The specific details of the above embodiment have been described in detail in the foregoing break detection method, so this place will not be repeated.

[0118] Experimental verification

[0119] An audio signal is input into an electronic device with the break detection method of the present application, and break detection is performed. The detection result is shown in FIG. 5 The dashed box in the figure shows the detected break part. At the beginning and end of the audio signal, it is explained that this part has a perceptual break. At the same time, it can be seen that the detected break part appears for a long time. Clipping occurs, but the middle part of the audio signal also has clipping phenomenon, but the break is not detected by the method of the present application. This part belongs to the part with clipping but no perceptual break. If the existing break detection method based on clipping is directly used, it is obvious that the middle part will be judged to have a break. The experiment shows that the break detection accuracy of the method of the present application is higher.

[0120] Although most of the distortion is caused by signal clipping, clipping is not equivalent to distortion, because only the clipping lasting for a certain time will form the audible distortion. Based on this, in one aspect, the clipping confidence and the clipping proportion of the detection window are extracted, and the multiple sets of clipping features of the signal frame (detection window) lasting for a certain time are combined and analyzed, so that the false detection of the audio segment with clipping but without distortion is effectively avoided. On the other hand, the frequency energy feature is added to deal with the scene where the signal is not clipped but has large energy and audible distortion, so that the missed detection of this scene is effectively avoided. Specifically, considering that the signal frame with large energy center is a common feature of clipping and this special scene, the special scene is detected based on the energy center, and the missed detection rate is reduced.

[0121] The clipping confidence is determined by the statistical signal distribution, which avoids the setting of the clipping threshold, avoids the problem of clipping detection failure caused by unreasonable threshold selection, and determines the clipping confidence through the time domain signal frame, which greatly reduces the computational complexity compared with the frequency domain determination process. At the same time, the determined clipping confidence can be continuously distributed between 0 and 1, which can provide greater flexibility for the subsequent distortion detection strategy compared with the traditional two-part case.

[0122] Exemplary apparatus

[0123] It should be noted that the distortion detection method provided by the embodiments of the present disclosure can be executed by the corresponding device. Next, reference is made to FIG. 6 First, the distortion detection device of the exemplary embodiment of the present disclosure is described.

[0124] FIG. 6 A block diagram of a distortion detection device according to one embodiment of the present disclosure is schematically shown.

[0125] Referring to FIG. 6 The distortion detection device 600 according to one embodiment of the present disclosure can include a signal division module 610, a first determination module 620, a second determination module 630, and a distortion determination module 640, as shown in the figure. The signal division module 610 can be configured to divide the audio signal to be detected into N time domain signal frames, where N is a positive integer. The first determination module 620 can be configured to determine the first clipping confidence corresponding to each time domain signal frame according to the statistical distribution of the signal amplitude in the frame, and the first clipping confidence is used to distinguish between normal audio signals and clipped audio signals. The second determination module 630 can be configured to determine the second clipping confidence corresponding to the detection window formed by the continuous M time domain signal frames according to the first clipping confidence of the continuous M time domain signal frames, where M is a positive integer less than N. The distortion determination module 640 can be configured to determine that there is distortion in the corresponding detection window in response to the second clipping confidence being greater than a preset first threshold.

[0126] In some embodiments of the present disclosure, based on the foregoing scheme, the first determination module 620 comprises: a histogram construction module configured to, for each frame of the time domain signal frame, count the number of signals in each amplitude section to construct a statistical histogram; the amplitude section is obtained by dividing the amplitude interval formed by the maximum amplitude and the minimum amplitude of the time domain signal frame; a search module configured to search for target blocks in the statistical histogram, the target block being a block with both ends higher than the middle of the block, and the back end of the block higher than the front end of the block, the front end and the back end of the block being determined according to the order of searching; a distance determination module configured to determine the horizontal distance of both ends of each target block; a confidence determination module configured to determine the first clipping confidence according to the ratio of the maximum value in each horizontal distance to the total number of amplitude sections.

[0127] In some embodiments of the present disclosure, based on the foregoing scheme, the search module is further configured to: start from both ends of the statistical histogram respectively, and move to the middle in turn to search for the target block.

[0128] In some embodiments of the present disclosure, based on the foregoing scheme, the apparatus 600 further comprises a clipping ratio determination module configured to: for each frame of the time domain signal frame, determine the first clipping ratio of the frame according to the proportion of the target signal in the frame of the time domain signal frame, the target signal being the signal corresponding to the maximum amplitude and the minimum amplitude; determine the second clipping ratio of the corresponding detection window according to the first clipping ratios of the continuous M frames of the time domain signal; the break sound determination module 640 is further configured to: in response to the second clipping confidence being greater than a preset first threshold and the second clipping ratio being greater than a preset second threshold, determine that there is a break sound in the detection window.

[0129] In some embodiments of the present disclosure, based on the foregoing scheme, the apparatus 600 further comprises: a frequency domain feature determination module configured to determine the frequency domain energy feature of the time domain signal frame in the detection window; the break sound determination module 640 is further configured to: in response to the signal frame in the detection window not satisfying at least one of the following: the second clipping confidence is greater than a preset first threshold, and the second clipping ratio is greater than a preset second threshold, determine whether there is a break sound in the detection window according to the frequency domain energy feature.

[0130] In some embodiments of the present disclosure, based on the foregoing scheme, the frequency domain energy feature comprises a cutoff frequency value, and the frequency domain feature determination module is further configured to: perform time-frequency transformation on each frame of the time domain signal frame in the detection window to obtain a corresponding frequency domain signal; determine the energy center of the frequency domain signal as the cutoff frequency value of the signal frame.

[0131] In some embodiments of the present disclosure, based on the foregoing scheme, the break tone determination module 640 is further configured to: in response to the cutoff frequency value of the signal frame being greater than a frequency threshold value, determine that the signal frame has clipping, the frequency threshold value being determined based on the maximum frequency value of the frequency domain signal; and in response to there being at least a preset number of signal frames with clipping in the detection window, determine that the detection window has a break tone.

[0132] In some embodiments of the present disclosure, based on the foregoing scheme, the second determination module 630 is further configured to: perform first weighting processing on the first clipping confidence of the M time domain signal frames in the detection window to obtain a second clipping confidence; and the clipping ratio determination module is further configured to: perform second weighting processing on the first clipping ratio of the M time domain signal frames in the detection window to obtain a second clipping ratio.

[0133] In some embodiments of the present disclosure, based on the foregoing scheme, M is determined based on the minimum time length of a break tone formed by continuous clipping and the length of the time domain signal frame.

[0134] The specific details of each module or unit in the above break tone detection apparatus have been described in detail in the corresponding break tone detection method, and thus will not be described here again.

[0135] Exemplary media

[0136] After introducing the method of the example embodiments of the present application, next, the medium of the example embodiments of the present application is described.

[0137] In some possible embodiments, each aspect of the present application can also be implemented as a storage medium having program code stored thereon, the program code being used to implement the steps in the break tone detection method according to various example embodiments of the present application described in the “example method” part of the present specification when the program code is executed by a processor of a device.

[0138] Specifically, the processor of the device executes the program code to implement the following steps:

[0139] divide the audio signal to be detected into N time domain signal frames, N being a positive integer; determine a first clipping confidence corresponding to each time domain signal frame according to the statistical distribution of the signal amplitude in the frame, the first clipping confidence being used to distinguish between normal audio signals and clipping audio signals; determine a second clipping confidence corresponding to a detection window formed by continuous M time domain signal frames according to the first clipping confidence of the continuous M time domain signal frames; M being a positive integer smaller than N; and in response to the second clipping confidence being greater than a preset first threshold value, determine that there is a break tone in the corresponding detection window.

[0140] The above is a schematic solution of the computer readable storage medium of the embodiment. It should be noted that the technical solution of the storage medium and the technical solution of the above-described broken sound detection method belong to the same concept, and the details of the technical solution of the storage medium that are not described in detail can be referred to the description of the technical solution of the broken sound detection method.

[0141] It should be noted that the above storage medium can be a readable storage medium. The readable storage medium may, for example, but is not limited to, an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, device or component, or any combination thereof. More specific examples (non-exhaustive list) of the readable storage medium include an electrical connection having one or more wires, a portable disc, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber, a portable compact disc read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination of the above.

[0142] The program code contained in the readable storage medium can be transmitted by any suitable medium, including but not limited to wireless, wired, optical cable, RF, etc., or any suitable combination thereof.

[0143] The program code for performing the operations of the present application can be written in any combination of one or more programming languages, including an object-oriented programming language such as Java, C++, etc., and a conventional procedural programming language such as "C" language or similar programming languages. The program code can be executed entirely on the user electronic device, partially on the user electronic device and partially on a remote electronic device, or entirely on a remote electronic device or server. In the case of a remote electronic device, the remote electronic device can be connected to the user electronic device through any kind of network, including a local area network (LAN) or a wide area network (WAN), or can be connected to an external electronic device (for example, through the Internet by using an Internet service provider).

[0144] Exemplary electronic devices

[0145] After introducing the method, medium and device of the exemplary embodiment of the present disclosure, next, an electronic device according to another exemplary embodiment of the present disclosure is introduced.

[0146] Those skilled in the art can understand that each aspect of the present application can be implemented as a system, a method or a program product. Therefore, each aspect of the present application can be embodied in a form of entirely hardware, entirely software (including firmware, microcode, etc.), or a combination of hardware and software, which can be collectively referred to as "circuitry", "module" or "system".

[0147] The electronic device 700 according to this embodiment of the present application will be described below with reference to FIG. 7 FIG. 7 The electronic device 700 is merely an example and should not impose any limitation on the function and scope of use of the embodiments of the present application.

[0148] As shown in FIG. 7 The electronic device 700 is in the form of a general electronic device. The components of the electronic device 700 can include, but are not limited to, the at least one processing unit 710 described above, the at least one storage unit 720 described above, and a bus 730 connecting different system components, including the storage unit 720 and the processing unit 710.

[0149] The storage unit stores program codes which can be executed by the processing unit 710, so that the processing unit 710 performs the steps according to various exemplary embodiments of the present application described in the "Exemplary Method" section of the present specification.

[0150] The storage unit 720 can include a readable medium in the form of a volatile storage unit, such as a random access memory (RAM) 7201 and / or a cache memory 7202, and can further include a read-only memory (ROM) 7203.

[0151] The storage unit 720 can further include a program / utility 7204 having a set of program modules 7205, including but not limited to an operating system, one or more application programs, other program modules, and program data, each of which or some combination thereof can include implementation of a network environment.

[0152] The bus 730 can represent one or more of several types of bus structures, including a storage unit bus or storage unit controller, a peripheral bus, a graphics acceleration port, a processing unit bus, or a local bus using any of a variety of bus architectures.

[0153] ​The electronic device 700 can also communicate with one or more external devices such as a keyboard or a pointing device, through I / O interface 750 and display unit 740. In general, the electronic device 700 will communicate with one or more other devices (not shown) in a network environment via communication interface 760. For example, the electronic device 700 can communicate with an application server on the Internet. The application server can comprise a web server, a file server, a network server, or the like. The application server can be operated by the same or a different entity than the electronic device 700. In these and other embodiments, the electronic device 700 can serve in a role of a client to the application server. In still other embodiments, the electronic device 700 can serve in a role of a server or peer to the application server, or in some other role. The application server and the electronic device 700 can communicate information between each other using a protocol. The protocol can be a network protocol, a file transfer protocol, a wireless communication protocol, or the like. The protocol can be a custom protocol or a standard protocol, such as the Internet Protocol Suite, the Hypertext Transfer Protocol, the File Transfer Protocol, the Wireless Application Protocol, or the like.

[0154] Through the above description of the embodiments, those skilled in the art can easily understand that the example embodiments described herein can be implemented by software, or by software in combination with necessary hardware. Therefore, the technical solutions according to the embodiments of the present disclosure can be embodied in the form of a software product, which can be stored in a non-volatile storage medium (which can be a CD-ROM, a USB flash disk, a mobile hard disk, or the like) or a network, and includes a number of instructions to make an electronic device (which can be a personal computer, a server, a terminal device, or a network device, etc.) execute the methods according to the embodiments of the present disclosure.

[0155] The above is a schematic solution of the electronic device 700 of the present embodiment. It should be noted that the technical solution of the electronic device 700 belongs to the same concept as the technical solution of the broken note detection method described above, and the details of the technical solution of the electronic device which are not described in detail can be referred to the description of the technical solution of the broken note detection method.

[0156] It should be noted that although several modules or sub-modules of the broken note detection apparatus are mentioned in the above detailed description, such division is merely exemplary and not mandatory. In fact, according to the embodiments of the present disclosure, the features and functions of two or more modules or units described above can be embodied in one module or unit. Conversely, the features and functions of one module or unit described above can be further divided into a plurality of modules or units.

[0157] Furthermore, although the operations of the method of the present application are described in a particular, sequential order, this order is not meant to be a limitation and is not intended to imply that there is an absolute requirement that the operations be performed in the order described. On the contrary, many of the operations can be performed in any order, added to, or omitted from the operations described. Additionally or alternatively, certain steps can be performed in parallel with one another.

[0158] While the principles and spirit of the application have been described with reference to several specific embodiments, it is to be understood that the application is not limited to the specific embodiments disclosed and that modifications to, and the use of, each of the various aspects are intended to be included as is within the spirit and scope of the application. The application is intended to cover various modifications and equivalent arrangements included within the spirit and scope of the appended claims.

Claims

1. A method of detecting a break-up, characterized by The method comprises: dividing a to-be-detected audio signal into N time-domain signal frames, N being a positive integer; determining a first clipping confidence degree corresponding to each time-domain signal frame according to a statistical distribution of signal amplitudes in the time-domain signal frame, the first clipping confidence degree being used to distinguish between normal audio signals and clipped audio signals; determining a second clipping confidence degree corresponding to a detection window formed by M continuous time-domain signal frames according to the first clipping confidence degrees of the M continuous time-domain signal frames, M being a positive integer smaller than N; in response to the second clipping confidence degree being greater than a preset first threshold, determining that there is a clipped note in the corresponding detection window; wherein the determining of the first clipping confidence degree according to the statistical distribution of signal amplitudes in each time-domain signal frame comprises: for each time-domain signal frame, counting the number of signals in each amplitude section to construct a statistical histogram, the amplitude section being obtained by dividing an amplitude interval formed by the maximum amplitude and the minimum amplitude of the time-domain signal frame; finding a target block in the statistical histogram, the target block being a block in which the two ends are higher than the middle and the back end is higher than the front end, the front end and the back end being determined according to the order of the finding; determining the horizontal distance of the two ends of each target block; determining the first clipping confidence degree according to the ratio of the maximum value in each horizontal distance to the total number of amplitude sections.

2. The method of claim 1, wherein, The finding of the target block comprises: starting from the two ends of the statistical histogram respectively, moving to the middle in turn to find the target block.

3. The method of claim 1, wherein, The method further comprises: for each time-domain signal frame, determining a first clipping ratio of the time-domain signal frame according to the proportion of a target signal in the time-domain signal frame, the target signal being a signal corresponding to the maximum amplitude and the minimum amplitude; determining a second clipping ratio of the corresponding detection window according to the first clipping ratios of M continuous time-domain signal frames; the determination of the presence of a clipped note in the corresponding detection window comprises: in response to the second clipping confidence degree being greater than a preset first threshold and the second clipping ratio being greater than a preset second threshold, determining that there is a clipped note in the detection window.

4. The method of claim 3, wherein, The method further comprises: determining a frequency domain energy feature of the time-domain signal frames in the detection window; the determination of the presence of a clipped note in the corresponding detection window comprises: in response to the signal frames in the detection window not satisfying at least one of the following conditions: the second clipping confidence degree is greater than a preset first threshold, the second clipping ratio is greater than a preset second threshold, determining whether there is a clipped note in the detection window according to the frequency domain energy feature.

5. The method of claim 4, wherein, The frequency domain energy feature comprises a cutoff frequency value, and the determination of the frequency domain energy feature of the time-domain signal frames in the detection window comprises: performing time-frequency transformation on each time-domain signal frame in the detection window to obtain a corresponding frequency domain signal; determining the energy center of the frequency domain signal as the cutoff frequency value of the signal frame.

6. The method of claim 5, wherein, the determination of whether there is a clipped note in the detection window according to the frequency domain energy feature comprises: in response to the cutoff frequency value of the signal frame being greater than a frequency threshold value, determining that there is a clipped note in the signal frame, the frequency threshold value being determined based on the maximum frequency value of the frequency domain signal. In response to the detection window having at least a preset number of signal frames with clipping, it is determined that the detection window has a break tone.

7. The method of claim 3, wherein, The second clipping confidence of the detection window is determined according to the first clipping confidence of the continuous M time domain signal frames. The second clipping confidence is obtained by performing first weighting processing on the first clipping confidence of the M time domain signal frames in the detection window. The second clipping proportion of the detection window is determined according to the first clipping proportion of the continuous M time domain signal frames. The second clipping proportion is obtained by performing second weighting processing on the first clipping proportion of the M time domain signal frames in the detection window.

8. The method according to any one of claims 1 to 7, characterized in that, The M is determined based on the minimum time length of the continuous clipping forming a break tone and the length of the time domain signal frame.

9. An unvoiced sound detection device characterized by comprising: The device comprises: A signal division module configured to divide the audio signal to be detected into N time domain signal frames, the N being a positive integer; A first determination module configured to determine, according to the statistical distribution of the signal amplitude in each time domain signal frame, the first clipping confidence corresponding to the frame, the first clipping confidence being used to distinguish between normal audio signals and clipped audio signals; A second determination module configured to determine, according to the first clipping confidence of the continuous M time domain signal frames, the second clipping confidence corresponding to the detection window formed by the continuous M time domain signal frames, the M being a positive integer smaller than the N; A break tone determination module configured to determine, in response to the second clipping confidence being greater than a preset first threshold, that there is a break tone in the corresponding detection window; The first determination module comprises: A histogram construction module configured to, for each time domain signal frame, count the number of signals in each amplitude section to construct a statistical histogram, the amplitude section being obtained by dividing the amplitude interval formed by the maximum amplitude and the minimum amplitude of the time domain signal frame; A lookup module configured to look up target blocks in the statistical histogram, the target block being a block in which both ends are higher than the middle, and the back end is higher than the front end, the front end and the back end being determined according to the order of the lookup; A distance determination module configured to determine the horizontal distance between the two ends of each target block; A confidence determination module configured to determine the first clipping confidence according to the ratio of the maximum value in each horizontal distance to the total number of amplitude sections.

10. The apparatus of claim 9, wherein, The lookup module is further configured to: Respectively starting from both ends of the statistical histogram, sequentially moving to the middle to find the target block.

11. The apparatus of claim 9, wherein, The device further comprises a clipping proportion determination module configured to: For each time domain signal frame, determine the first clipping proportion of the frame according to the proportion of the target signal in the frame, the target signal being the signal corresponding to the maximum amplitude and the minimum amplitude; Determine the second clipping proportion of the detection window according to the first clipping proportion of the continuous M time domain signal frames; The break tone determination module is further configured to: In response to the second clipping confidence being greater than a preset first threshold and the second clipping proportion being greater than a preset second threshold, it is determined that there is a break tone in the detection window.

12. The apparatus of claim 11, wherein, The device further comprises: The frequency domain feature determination module is configured to determine a frequency domain energy feature of the time domain signal frame in the detection window. The clip determination module is further configured to: in response to the signal frame in the detection window not satisfying at least one of the following conditions: the second clip confidence is greater than a preset first threshold value, the second clip proportion is greater than a preset second threshold value, determine, according to the frequency domain energy feature, whether there is a clip in the detection window.

13. The apparatus of claim 12, wherein, The frequency domain energy feature includes a cutoff frequency value, and the frequency domain feature determination module is further configured to: perform time-frequency transformation on each time domain signal frame in the detection window to obtain a corresponding frequency domain signal; determine an energy center of the frequency domain signal as the cutoff frequency value of the signal frame.

14. The apparatus of claim 13, wherein, The clip determination module is further configured to: in response to the cutoff frequency value of the signal frame being greater than a frequency threshold value, determine that the signal frame has a clip, the frequency threshold value being determined based on a maximum frequency value of the frequency domain signal; in response to there being at least a preset number of signal frames with clips in the detection window, determine that there is a clip in the detection window.

15. The apparatus of claim 11, wherein, The second determination module is further configured to: perform first weighting processing on the first clip confidence of the M time domain signal frames in the detection window to obtain the second clip confidence; The clip proportion determination module is further configured to: perform second weighting processing on the first clip proportion of the M time domain signal frames in the detection window to obtain the second clip proportion.

16. The apparatus of any one of claims 9-15, wherein, The M is determined based on a minimum time length of a clip forming a clip and a length of a time domain signal frame.

17. An electronic device, comprising: comprising: a processor and a memory, the memory storing executable instructions, and the processor being configured to invoke the executable instructions stored in the memory to execute the method according to any one of claims 1 to 8.

18. A storage medium, characterized by a program stored thereon, the program being executed by a processor to implement the method according to any one of claims 1 to 8.

Images