EP3089163A1 - Methode zur niedrigverlust-entfernung von stationären und nichtstationären kurzzeit-interferenzen - Google Patents

Methode zur niedrigverlust-entfernung von stationären und nichtstationären kurzzeit-interferenzen Download PDF

Info

Publication number
EP3089163A1
EP3089163A1 EP16167048.4A EP16167048A EP3089163A1 EP 3089163 A1 EP3089163 A1 EP 3089163A1 EP 16167048 A EP16167048 A EP 16167048A EP 3089163 A1 EP3089163 A1 EP 3089163A1
Authority
EP
European Patent Office
Prior art keywords
signal
input
noise
windows
dominant
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Granted
Application number
EP16167048.4A
Other languages
English (en)
French (fr)
Other versions
EP3089163B1 (de
Inventor
Tilman Herberger
Titus Tost
Georg Flemming
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Bellevue Investments GmbH and Co KGaA
Original Assignee
Bellevue Investments GmbH and Co KGaA
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Family has litigation
First worldwide family litigation filed litigation Critical https://patents.darts-ip.com/?family=55952965&utm_source=google_patent&utm_medium=platform_link&utm_campaign=public_patent_search&patent=EP3089163(A1) "Global patent litigation dataset” by Darts-ip is licensed under a Creative Commons Attribution 4.0 International License.
Priority claimed from US14/702,541 external-priority patent/US9552829B2/en
Application filed by Bellevue Investments GmbH and Co KGaA filed Critical Bellevue Investments GmbH and Co KGaA
Publication of EP3089163A1 publication Critical patent/EP3089163A1/de
Application granted granted Critical
Publication of EP3089163B1 publication Critical patent/EP3089163B1/de
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • G—PHYSICS
    • G10—MUSICAL INSTRUMENTS; ACOUSTICS
    • G10L—SPEECH ANALYSIS OR SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING; SPEECH OR AUDIO CODING OR DECODING
    • G10L21/00—Processing of the speech or voice signal to produce another audible or non-audible signal, e.g. visual or tactile, in order to modify its quality or its intelligibility
    • G10L21/02—Speech enhancement, e.g. noise reduction or echo cancellation
    • G10L21/0208—Noise filtering
    • G—PHYSICS
    • G10—MUSICAL INSTRUMENTS; ACOUSTICS
    • G10L—SPEECH ANALYSIS OR SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING; SPEECH OR AUDIO CODING OR DECODING
    • G10L21/00—Processing of the speech or voice signal to produce another audible or non-audible signal, e.g. visual or tactile, in order to modify its quality or its intelligibility
    • G10L21/02—Speech enhancement, e.g. noise reduction or echo cancellation
    • G10L21/0208—Noise filtering
    • G10L21/0216—Noise filtering characterised by the method used for estimating noise
    • G10L21/0232—Processing in the frequency domain
    • G—PHYSICS
    • G10—MUSICAL INSTRUMENTS; ACOUSTICS
    • G10L—SPEECH ANALYSIS OR SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING; SPEECH OR AUDIO CODING OR DECODING
    • G10L21/00—Processing of the speech or voice signal to produce another audible or non-audible signal, e.g. visual or tactile, in order to modify its quality or its intelligibility
    • G10L21/02—Speech enhancement, e.g. noise reduction or echo cancellation
    • G10L21/0208—Noise filtering
    • G10L2021/02085—Periodic noise

Definitions

  • the present invention relates to the general subject matter of creating and analyzing video works and, more specifically, to systems and methods for analyzing the audio part of video recordings to attenuate ambient noise.
  • Removal or attenuation of ambient noise in video recordings is a subject matter in which a number of different approaches exist, all trying to be the most effective solution without harming the integrity of the input signal.
  • Many known methods to attenuate or remove ambient noise in video recordings at the current state of the art utilize the principle of "spectral subtraction".
  • a filter function is determined by estimating the average magnitude of the interfering signal and subtracting it from the magnitude of the target signal.
  • the characteristics of the interfering signal can be automatically determined by a targeted search for unwanted components which might be found, for example, in breaks between the speech (or other target) content with the audio material.
  • Another way to determine an estimate of the interfering signal is to let the user define a noise sample - a sample of audio material that contains only or predominantly the interfering signal. This approach is well-known and is often utilized in software based solutions.
  • the approaches known in the state of the art typically operate by determining the average of short time magnitudes of the interfering signal. This averaging is useful for interfering signals with a stochastic character such as ambient noise or for stationary interfering signals such as a buzzing-type signal.
  • Interfering signals from motors, especially zoom- and autofocus motors in digital cameras usually exhibit a non-stationary behavior: The sound behavior of the interference can be variable and its frequency response often changes fluently / continuously.
  • the pitch will generally depend on the intended focal length, respectively on the focal length of the objective that has been passed through during the set up process. In this case the averaging of the short time magnitude of the interfering signal would lead to a broad band filtering which would unnecessarily remove substantial parts of the input signal.
  • Methods according to the state of the art usually implement the filtering process continuously for the whole length of the impaired input signal.
  • the input signal will also be subjected to a filtering process in the undisturbed passages that are often longer, ultimately leading to a deterioration of the audio quality of the input signal in these undisturbed passages.
  • An embodiment removes or attenuates temporarily occurring (e.g., transient) stationary and non-stationary interfering noise, e.g., interference from zoom and autofocus motors in digital video cameras or digital photo cameras with video recording functionality.
  • An embodiment may be applied only to the particular passages containing interfering noise.
  • One embodiment modifies the process of "spectral subtraction" by incorporating a dynamic approach, with the goal being to limit the filtering electively to only the disturbed or otherwise noise contaminated portions of the signal.
  • an embodiment makes it possible to parameterize the algorithm in such a way that the transfer function is primarily determined by the spectral matching maxima of the short time magnitudes of the input signal and the interfering signal. In this case the relevant tonal parts of the interference will be removed or attenuated and the rest of the input signal will be only minimally affected.
  • One object of an embodiment is to modify the standard process of "spectral subtraction" by utilizing the short time magnitude of the interfering signal that has the highest conformity with the determined short time magnitude of the input signal containing the interference. This correspondence is then utilized for the calculation of the transfer function of the filter, thereby attenuating the interfering signal. With this approach it will be possible for the resulting transfer function to stay comparatively narrow-banded, which will impair the input signal much less.
  • the instant invention will be explained with reference to four stages that together make up the algorithm for low-loss removal of stationary and non-stationary short time interference.
  • This approach is robust and tolerant of the volume differences between the impaired input signal and the recorded noise sample of the interference signal.
  • An embodiment begins with an analysis of the input signal, Component A.
  • This phase is intended to detect the best possible match between an example of an interfering signal, preferably a sample in which the target signal is absent so that noise / interfering signal alone is present, with the portions of the input signal that is contaminated by the same sort of noise and, thus, that contains signal plus interference.
  • the second Component B is performs an analysis of the input signal to detect adjacent sections containing evidence of interference.
  • the third Component C will remove the interfering signal via adaptive filtering.
  • the transfer function of the adaptive filter will be determined.
  • calculation of the removal of the interference signal and calculation of the associated transfer function could be implemented in real time, although that is not a requirement.
  • the input file or recording containing a noise sample of the interfering signal and the file or recording containing the impaired input signal will be analyzed and the results stored in memory.
  • the results of the Component A analysis will be passed to Component B, whose results will also be stored in memory.
  • the resulting data from Components A and B will then be utilized in the calculation of the transfer function of the adaptive noise attenuating filter generated in Component C handles removal of the interfering noise, preferably in real time.
  • the data from Component A will be used to determine which section of the noise sample should be used for the calculation of the transfer function of the interference noise removing filter.
  • the data from Component B will be utilized to determine whether or not dampening of the interference noise is carried out section by section in full or lesser strength. The level of dampening will be fully adjustable by the user.
  • At least a portion of the instant invention could be implemented in form of software running on a user's computer 600, where the term "computer” is intended to encompass any programmable device capable of performing signal processing of the sort described herein (including, without limitation, cell phones 650, tablet computers 640, etc.).
  • a computer will have some amount of program memory and data storage capability 620 (whether internal, external, or accessible via a network) as is conventionally utilized by such units.
  • an external camera 610 of some sort would be utilized with - and will preferably be connectible directly or indirectly (e.g., via removable storage) to -the computer 600 so that video and audio and/or graphic information can be transferred to and from the computer.
  • the camera 610 will be a digital video camera with audio recording capabilities, although that is not a requirement and in some instances the input might be a recording obtained via one or more microphones 630 which could be connected directly to the computer or connectable indirectly via removable storage or wireless communication.
  • the camera and/or microphone might be integrated into an electronic device rather than being a stand alone component.
  • the camera will preferably be digital in nature, any sort of camera might be used, provided that the proper interface between it and the computer is utilized.
  • An embodiment preferably begins with the reading of two digital signals 105 and 125 into a personal or other computer, to include portable computers such as tablet computers, cell phones, etc.
  • the input to box 105 is the input or target signal, e.g., the audio part of a video recording
  • the input signal 125 is a noise-only recording or, more specifically, a digital representation of the noise that contaminates the target recording 105 .
  • the selection of the noise-only recording will be performed at the hands of the user, who might be instructed to record a sample of the noise, select a passage within the target signal that is essentially all noise, etc.
  • the first few steps of the current embodiment analyze the input to find the best possible matches of a noise sample of the interference signal 125 with the sections of the impaired (i.e., noise plus signal) input signal 105.
  • the best matches between the single sections of the recorded interference signal - the noise sample and the single sections of the impaired input signal will be determined in this embodiment by matching the frequencies of spectral maxima - the peaks.
  • the number of spectral maxima (peaks) with matching frequency can be utilized to determine the quality of the match.
  • both data files 105 and 125 will be partitioned into overlapping windows or, equivalently, blocks 110 and 130 which are weighted by a windowing function (e.g., a Gaussian taper, a Hanning taper, a Hamming taper, etc.).
  • a windowing function e.g., a Gaussian taper, a Hanning taper, a Hamming taper, etc.
  • the amount of overlap between adjacent windows might be 75%, but other choices are certainly possible.
  • FFT Fast Fourier Transform
  • the absolute value the spectral short time magnitudes of each of the windows will be calculated 115 and 135 .
  • a window length of 4096 samples might be used.
  • the sample rate of the input can be freely chosen but often it will be between about 16 and 192 kHz. One sample rate that is often useful is 44.1 or 48 kHz.
  • the short time magnitudes will be analyzed to identify their spectral maxima (peaks) 120 and 140 , each of which corresponds to a dominant frequency that is present in the signal.
  • the phrase "dominant frequency" will be used to indicate those frequencies in either the target signal or the noise signal that correspond to significant peaks in the frequency spectrum and, more particularly, to those frequencies that correspond to the largest peaks in the spectrum.
  • the spectral maxima of the short time magnitudes of the impaired input signal will be compared with the spectral maxima of the short time magnitudes of the noise sample to determine where there is a match of the frequencies in both signals 145 .
  • peaks in the frequency spectrum will be identified as dominant.
  • the most significant peaks will be used in the steps that follow.
  • the number of peaks identified might be between 10 and 50.
  • the individual short time magnitudes and the positions of the spectral maxima, the peaks of the noise sample will be identified and stored at least temporarily, because the locations of the peaks is used in the calculation of the transfer function of the adaptive interference signal removing filter as discussed below.
  • the signals will be analyzed to detect portions of the target signal that are contiguous to sections that have already been flagged as containing interfering noise 155 .
  • One aim of this step is to utilize the results from component A to determine the sections of the noise contaminated input signal in which the impairing noise is present in windows other than those already identified.
  • one reason for performing this step is that if a block is determined to be contaminated by noise it will often be the case that blocks adjacent to it (either earlier or later in time) will also be contaminated by noise.
  • the analysis will be performed at least at each end of a group of blocks.
  • groups of 10 contiguous blocks will be identified and used to determine the value with the minimal match of spectral maxima, peaks. More specifically, for each group of 10 blocks, the average of the maximum peak values in each block and the average of the minimum peak values in each block will be calculated.
  • a threshold value might be selected which sets a minimum as to the spectral matches that will used subsequently: e.g., spectral peaks below this value will not be counted.
  • the sensitivity of the detection will be adjustable with a threshold factor (e.g., a multiplier) Parameter P1, which can be selected by the user 150 .
  • P1 might be based on some measure of the variability of the minima so that peaks that exceed P1 will be those peaks that are statistically higher than the average (e.g., one standard deviation higher).
  • a P1 might also be used in connection with an analysis of the number of matching peaks between the noise signal and the subject signal as modified by thresholding or other screening.
  • counts of the number of matching peaks will be used instead. That is, this variation the number of matching peaks within each block will be determined. The average number of matches in all blocks will then be calculated. Then, those blocks that have an above average number of matches will be flagged as having noise present therein.
  • a numerical threshold might be imposed so that only those counts that exceed the average by some designated amount (e.g., by some percentage of the average, by a multiple of the standard deviation of the counts in that group, etc.) would be designated as being contaminated by noise.
  • the neighboring (e.g., contiguous) blocks of the signal blocks in the target that are designated as impaired will be inspected to determine if the number of matching spectral maxima are slightly (e.g., 10%) above the average value of the average count of spectral matches in a block.
  • the signal block will also be identified as an impaired signal block.
  • the same process could then be applied to the corresponding signal section of the other stereo channel, if such is available.
  • the signal block of the other stereo channel could also be assigned this property, provided that the determined number of matching spectral maxima is slightly above the average value for the minimum spectral maxima matching.
  • a noise source might be present on one channel but not the other or that the noise characteristics on two channels might be different. In such an instance, the two channels might be processed separately according to the methods taught herein.
  • the input signal 200 will be partitioned into overlapping blocks 205, a windowing function will be utilized as described previously, and a FFT (or other transform) will be used to transform the signal into the frequency domain 210, thereby producing a complex spectrum.
  • a FFT or other transform
  • the resulting complex spectrum of the input will be multiplied by a short time invariant filter function that is designed to remove or attenuate the noise relative to the target signal.
  • the next step according to this embodiment will transform the filtered spectrum back into the time domain via an inverse FFT, followed by a windowing operation 220 and a summation or other compilation of the individual time-domain blocks 225 to construct a non-windowed and filtered signal.
  • the transfer function of the filter will be subjected to temporal smoothing to reduce artifacts (chirping artifacts, for example) 240 before application to the target signal.
  • the time constants used the control the decline and rise of the smoothing filter can be defined separately in the some embodiments.
  • the input signal will next be mixed with the filtered signal. If in component B a section is determined to be undisturbed 245, the mixing ratio will be changed according to the setup of parameter P4 (which might be allowed to vary between 0.0 and 1.0) in such a way that the portion of the unfiltered input signal in the combination increases accordingly 250. In an embodiment it may not be necessary to apply the filtering process to the sections that have been flagged as undisturbed if the interference signal also contains noise and buzzing sounds which impair the input signal continuously.
  • FIGS. 3 and 4 illustrate an embodiment of a temporal curve of the sort used to detect contiguous impaired sections in an audio signal that contain speech content that is partially impaired.
  • the curve 310 depicts the number of matching spectral maxima within each signal block.
  • Line 320 represents the average value for the minimum of the contiguous spectral maxima.
  • the vertical axis is the number of matching spectral peaks between the target signal and the noise signal in each window.
  • the horizontal axis is the window number within the target signal.
  • Figure 4 illustrates an embodiment of the previous analysis and detection step in a subsequent stage wherein some windows have been categorized as being impaired.
  • the curve 410 is the same as curve 310 and depicts the number of matching spectral maxima for each signal block.
  • the regions that are delimited by the line 420 correspond to sections of the target signal that have been identified as impaired sections.
  • the transfer function of the adaptive filter will be determined.
  • the calculation of the transfer function is implemented via the process of spectral subtraction (as that term is known and used by those of ordinary skill in the art) using the results of the analysis of the matching spectral maxima (peaks) of the short time magnitude of the impaired input signal and the short time magnitude of the noise signal with highest matching to the impaired signal block.
  • the short time magnitudes and the information about the existing spectral maxima of the interference signal are provided by the results of the analysis carried out by Component A.
  • a separately calculated adaptive filter might be applied to each block, to each groups of blocks (e.g., 10 blocks), or a single calculated adaptive filter might be applied to the entire input signal.
  • the short time magnitude of the noise impaired signal can be determined by implementing the following steps as set out in the example of Figure 5 .
  • a digitized audio input signal containing noise-only (or predominantly noise) will be accessed 500 , e.g., read from disk, read in real time from a microphone, extracted from a video clip, etc. That signal will then be partitioned into blocks 505 , windowed 510 and transformed to the frequency domain 515 (preferably via a FFT).
  • the Component A results will be provided (box 454 ) as indicated in Figure 5 which might include the short time magnitude of the interfering signal that has the highest match to the impaired signal block (box 550 ) as well as the spectral maxima of the peaks with the highest match to the impaired signal block (box 555 ).
  • the process of spectral subtraction will be used to determine the transfer function by subtracting the short time magnitude of the interference signal from the associated short time magnitude of the impaired input signal (box 535 ) according to methods well know to those of ordinary skill in the art.
  • the Parameter P2 has been provided so that a user can adjust the amplitude of the spread of the impairment 525 (e.g., the range of spectral amplitudes that are considered to be dominant, e.g., amplitudes of frequencies that fall 10% of the maximum amplitude) and Parameter P3 can be utilized to adjust the bandwidth of the filter 530 by, for example, restricting the range of frequencies that are considered for matching purposes.
  • the transfer function can be set to the value zero 540, thereby filtering the target signal so as to attenuate the matching frequencies.
  • the value of the parameter P3 is chosen to be relatively low, the contribution of the spectral subtraction in the calculation of the transfer function (box 560 ) of the filter will be deemphasized.
  • the transfer function will be primarily defined by the matching peaks.
  • relevant frequency components of the noise signal will be removed from the target, which provides an advantage in that the damage to the target signal will be reduced.
  • this approach presents a sensible alternative to a complete removal of the interfering signal.
  • the filtered blocks will be reassembled to form a modified version of the target signal in which the noise signal has been attenuated.
  • the final product would then be available to be performed for a user.
  • the performance might take place using the speaker in a user's computer, as one specific example.
  • the instant invention might be particularly useful in processing the audio component of a video recording to improve the conversation (or other signal component) therein. It could also be used to attenuate crowd noise during the recording of a live music concert or live speech, among many other uses. It could also be used to attenuate zoom motor noise in video recordings.
  • noise signal or an "impairment signal”
  • that term should be broadly construed to include instances where there might be some of the signal which it is desired to enhance relative to noise (e.g., speech) within the noise signal but where the noise signal predominantly contains contaminating noise which is it desired to deemphasize or remove from the target signal.
  • the noise signal might be conventional (e.g., white noise, 60 Hz noise, engine noise, etc.) or it might be, for example, music that is playing while an individual is talking.
  • the "noise signal” will be any component of the recorded audio other than the information which it is desired to enhance (e.g., speech).
  • the instant invention provides a substantial improvement for both novice and professional users when editing audio recordings and primarily for attenuating interference signals in speech signals of video recordings.
  • the instant invention requires minimal user interaction, no definition of multiple parameters, it is an automatic process that analyzes the input signal and incorporates specific processes to process the input signal and to remove interference signals without overly harming the input signal.
  • Methods of the present invention may be implemented by performing or completing manually, automatically, or a combination thereof, selected steps or tasks.
  • method may refer to manners, means, techniques and procedures for accomplishing a given task including, but not limited to, those manners, means, techniques and procedures either known to, or readily developed from known manners, means, techniques and procedures by practitioners ⁇ of the art to which the invention belongs.
  • the term "at least” followed by a number is used herein to denote the start of a range beginning with that number (which may be a ranger having an upper limit or no upper limit, depending on the variable being defined). For example, “at least 1” means 1 or more than 1.
  • the term “at most” followed by a number is used herein to denote the end of a range ending with that number (which may be a range having 1 or 0 as its lower limit, or a range having no lower limit, depending upon the variable being defined). For example, “at most 4" means 4 or less than 4, and "at most 40%” means 40% or less than 40%.
  • a range is given as "(a first number) to (a second number)" or "(a first number) - (a second number)"
  • 25 to 100 should be interpreted to mean a range whose lower limit is 25 and whose upper limit is 100.
  • every possible subrange or interval within that range is also specifically intended unless the context indicates to the contrary.
  • ranges for example, if the specification indicates a range of 25 to 100 such range is also intended to include subranges such as 26 -100, 27-100, etc., 25-99, 25-98, etc., as well as any other possible combination of lower and upper values within the stated range, e.g., 33-47, 60-97, 41-45, 28-96, etc.
  • integer range values have been used in this paragraph for purposes of illustration only and decimal and fractional values (e.g., 46.7 - 91.3) should also be understood to be intended as possible subrange endpoints unless specifically excluded.
  • the defined steps can be carried out in any order or simultaneously (except where context excludes that possibility), and the method can also include one or more other steps which are carried out before any of the defined steps, between two of the defined steps, or after all of the defined steps (except where context excludes that possibility).
  • the instant invention will provide an automatic mode, which automatically attenuates video recordings in video cameras, therewith providing video recordings with improved quality audio.
EP16167048.4A 2015-05-01 2016-04-26 Methode zur niedrigverlust-entfernung von stationären und nichtstationären kurzzeit-interferenzen Active EP3089163B1 (de)

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
US14/702,541 US9552829B2 (en) 2014-05-01 2015-05-01 System and method for low-loss removal of stationary and non-stationary short-time interferences

Publications (2)

Publication Number Publication Date
EP3089163A1 true EP3089163A1 (de) 2016-11-02
EP3089163B1 EP3089163B1 (de) 2017-07-05

Family

ID=55952965

Family Applications (1)

Application Number Title Priority Date Filing Date
EP16167048.4A Active EP3089163B1 (de) 2015-05-01 2016-04-26 Methode zur niedrigverlust-entfernung von stationären und nichtstationären kurzzeit-interferenzen

Country Status (1)

Country Link
EP (1) EP3089163B1 (de)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN106993097A (zh) * 2017-03-31 2017-07-28 维沃移动通信有限公司 一种音乐播放方法及移动终端

Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20040165736A1 (en) * 2003-02-21 2004-08-26 Phil Hetherington Method and apparatus for suppressing wind noise
US20090177466A1 (en) * 2007-12-20 2009-07-09 Kabushiki Kaisha Toshiba Detection of speech spectral peaks and speech recognition method and system

Patent Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20040165736A1 (en) * 2003-02-21 2004-08-26 Phil Hetherington Method and apparatus for suppressing wind noise
US20090177466A1 (en) * 2007-12-20 2009-07-09 Kabushiki Kaisha Toshiba Detection of speech spectral peaks and speech recognition method and system

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
VASEGHI'S: "Advanced Signal Processing and Noise Reduction", 2000, JOHN WILEY & SONS

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN106993097A (zh) * 2017-03-31 2017-07-28 维沃移动通信有限公司 一种音乐播放方法及移动终端

Also Published As

Publication number Publication date
EP3089163B1 (de) 2017-07-05

Similar Documents

Publication Publication Date Title
US10210883B2 (en) Signal processing apparatus for enhancing a voice component within a multi-channel audio signal
AU2002240461B2 (en) Comparing audio using characterizations based on auditory events
US7508948B2 (en) Reverberation removal
US8027743B1 (en) Adaptive noise reduction
CN110619882B (zh) 用于降低去相关器电路中瞬态信号的时间伪差的系统和方法
AU2006302549A1 (en) Neural network classifier for seperating audio sources from a monophonic audio signal
CN108305637B (zh) 耳机语音处理方法、终端设备及存储介质
US9552829B2 (en) System and method for low-loss removal of stationary and non-stationary short-time interferences
CN110277087B (zh) 一种广播信号预判预处理方法
JP5752324B2 (ja) 雑音の入った音声信号中のインパルス性干渉の単一チャネル抑制
Towsey Noise removal from wave-forms and spectrograms derived from natural recordings of the environment
KR960005740B1 (ko) 음성신호처리장치
CN110503967B (zh) 一种语音增强方法、装置、介质和设备
WO2017158102A1 (en) Apparatus and method for harmonic-percussive-residual sound separation using a structure tensor on spectrograms
JP2014513320A (ja) オーディオ信号におけるドミナント周波数を減衰する方法及び装置
EP3089163B1 (de) Methode zur niedrigverlust-entfernung von stationären und nichtstationären kurzzeit-interferenzen
JP2013205831A (ja) 音声品質客観評価装置及び方法
US9269370B2 (en) Adaptive speech filter for attenuation of ambient noise
US9978393B1 (en) System and method for automatically removing noise defects from sound recordings
Brandt et al. Automatic detection of hum in audio signals
US20230290367A1 (en) Hum noise detection and removal for speech and music recordings
RU2591012C2 (ru) Устройство и способ обработки переходных процессов для аудио сигналов с изменением скорости воспроизведения или высоты тона
EP3032536B1 (de) Adaptiver sprachfilter zur unterdrückung von umgebungsgeräuschen
JP2011217139A (ja) 信号処理装置および方法、並びにプログラム
von Zeddelmann A feature-based approach to noise robust speech detection

Legal Events

Date Code Title Description
PUAI Public reference made under article 153(3) epc to a published international application that has entered the european phase

Free format text: ORIGINAL CODE: 0009012

AK Designated contracting states

Kind code of ref document: A1

Designated state(s): AL AT BE BG CH CY CZ DE DK EE ES FI FR GB GR HR HU IE IS IT LI LT LU LV MC MK MT NL NO PL PT RO RS SE SI SK SM TR

AX Request for extension of the european patent

Extension state: BA ME

17P Request for examination filed

Effective date: 20161027

RBV Designated contracting states (corrected)

Designated state(s): AL AT BE BG CH CY CZ DE DK EE ES FI FR GB GR HR HU IE IS IT LI LT LU LV MC MK MT NL NO PL PT RO RS SE SI SK SM TR

GRAP Despatch of communication of intention to grant a patent

Free format text: ORIGINAL CODE: EPIDOSNIGR1

INTG Intention to grant announced

Effective date: 20170224

GRAS Grant fee paid

Free format text: ORIGINAL CODE: EPIDOSNIGR3

GRAA (expected) grant

Free format text: ORIGINAL CODE: 0009210

AK Designated contracting states

Kind code of ref document: B1

Designated state(s): AL AT BE BG CH CY CZ DE DK EE ES FI FR GB GR HR HU IE IS IT LI LT LU LV MC MK MT NL NO PL PT RO RS SE SI SK SM TR

REG Reference to a national code

Ref country code: GB

Ref legal event code: FG4D

REG Reference to a national code

Ref country code: CH

Ref legal event code: EP

REG Reference to a national code

Ref country code: AT

Ref legal event code: REF

Ref document number: 907090

Country of ref document: AT

Kind code of ref document: T

Effective date: 20170715

REG Reference to a national code

Ref country code: IE

Ref legal event code: FG4D

REG Reference to a national code

Ref country code: DE

Ref legal event code: R096

Ref document number: 602016000122

Country of ref document: DE

REG Reference to a national code

Ref country code: NL

Ref legal event code: MP

Effective date: 20170705

REG Reference to a national code

Ref country code: AT

Ref legal event code: MK05

Ref document number: 907090

Country of ref document: AT

Kind code of ref document: T

Effective date: 20170705

REG Reference to a national code

Ref country code: LT

Ref legal event code: MG4D

PG25 Lapsed in a contracting state [announced via postgrant information from national office to epo]

Ref country code: NL

Free format text: LAPSE BECAUSE OF FAILURE TO SUBMIT A TRANSLATION OF THE DESCRIPTION OR TO PAY THE FEE WITHIN THE PRESCRIBED TIME-LIMIT

Effective date: 20170705

PG25 Lapsed in a contracting state [announced via postgrant information from national office to epo]

Ref country code: FI

Free format text: LAPSE BECAUSE OF FAILURE TO SUBMIT A TRANSLATION OF THE DESCRIPTION OR TO PAY THE FEE WITHIN THE PRESCRIBED TIME-LIMIT

Effective date: 20170705

Ref country code: LT

Free format text: LAPSE BECAUSE OF FAILURE TO SUBMIT A TRANSLATION OF THE DESCRIPTION OR TO PAY THE FEE WITHIN THE PRESCRIBED TIME-LIMIT

Effective date: 20170705

Ref country code: AT

Free format text: LAPSE BECAUSE OF FAILURE TO SUBMIT A TRANSLATION OF THE DESCRIPTION OR TO PAY THE FEE WITHIN THE PRESCRIBED TIME-LIMIT

Effective date: 20170705

Ref country code: HR

Free format text: LAPSE BECAUSE OF FAILURE TO SUBMIT A TRANSLATION OF THE DESCRIPTION OR TO PAY THE FEE WITHIN THE PRESCRIBED TIME-LIMIT

Effective date: 20170705

Ref country code: SE

Free format text: LAPSE BECAUSE OF FAILURE TO SUBMIT A TRANSLATION OF THE DESCRIPTION OR TO PAY THE FEE WITHIN THE PRESCRIBED TIME-LIMIT

Effective date: 20170705

Ref country code: NO

Free format text: LAPSE BECAUSE OF FAILURE TO SUBMIT A TRANSLATION OF THE DESCRIPTION OR TO PAY THE FEE WITHIN THE PRESCRIBED TIME-LIMIT

Effective date: 20171005

REG Reference to a national code

Ref country code: NL

Ref legal event code: NE

Effective date: 20180123

PG25 Lapsed in a contracting state [announced via postgrant information from national office to epo]

Ref country code: LV

Free format text: LAPSE BECAUSE OF FAILURE TO SUBMIT A TRANSLATION OF THE DESCRIPTION OR TO PAY THE FEE WITHIN THE PRESCRIBED TIME-LIMIT

Effective date: 20170705

Ref country code: ES

Free format text: LAPSE BECAUSE OF FAILURE TO SUBMIT A TRANSLATION OF THE DESCRIPTION OR TO PAY THE FEE WITHIN THE PRESCRIBED TIME-LIMIT

Effective date: 20170705

Ref country code: PL

Free format text: LAPSE BECAUSE OF FAILURE TO SUBMIT A TRANSLATION OF THE DESCRIPTION OR TO PAY THE FEE WITHIN THE PRESCRIBED TIME-LIMIT

Effective date: 20170705

Ref country code: BG

Free format text: LAPSE BECAUSE OF FAILURE TO SUBMIT A TRANSLATION OF THE DESCRIPTION OR TO PAY THE FEE WITHIN THE PRESCRIBED TIME-LIMIT

Effective date: 20171005

Ref country code: RS

Free format text: LAPSE BECAUSE OF FAILURE TO SUBMIT A TRANSLATION OF THE DESCRIPTION OR TO PAY THE FEE WITHIN THE PRESCRIBED TIME-LIMIT

Effective date: 20170705

Ref country code: IS

Free format text: LAPSE BECAUSE OF FAILURE TO SUBMIT A TRANSLATION OF THE DESCRIPTION OR TO PAY THE FEE WITHIN THE PRESCRIBED TIME-LIMIT

Effective date: 20171105

Ref country code: GR

Free format text: LAPSE BECAUSE OF FAILURE TO SUBMIT A TRANSLATION OF THE DESCRIPTION OR TO PAY THE FEE WITHIN THE PRESCRIBED TIME-LIMIT

Effective date: 20171006

REG Reference to a national code

Ref country code: DE

Ref legal event code: R097

Ref document number: 602016000122

Country of ref document: DE

PG25 Lapsed in a contracting state [announced via postgrant information from national office to epo]

Ref country code: CZ

Free format text: LAPSE BECAUSE OF FAILURE TO SUBMIT A TRANSLATION OF THE DESCRIPTION OR TO PAY THE FEE WITHIN THE PRESCRIBED TIME-LIMIT

Effective date: 20170705

Ref country code: DK

Free format text: LAPSE BECAUSE OF FAILURE TO SUBMIT A TRANSLATION OF THE DESCRIPTION OR TO PAY THE FEE WITHIN THE PRESCRIBED TIME-LIMIT

Effective date: 20170705

PLBE No opposition filed within time limit

Free format text: ORIGINAL CODE: 0009261

STAA Information on the status of an ep patent application or granted ep patent

Free format text: STATUS: NO OPPOSITION FILED WITHIN TIME LIMIT

PG25 Lapsed in a contracting state [announced via postgrant information from national office to epo]

Ref country code: SM

Free format text: LAPSE BECAUSE OF FAILURE TO SUBMIT A TRANSLATION OF THE DESCRIPTION OR TO PAY THE FEE WITHIN THE PRESCRIBED TIME-LIMIT

Effective date: 20170705

Ref country code: IT

Free format text: LAPSE BECAUSE OF FAILURE TO SUBMIT A TRANSLATION OF THE DESCRIPTION OR TO PAY THE FEE WITHIN THE PRESCRIBED TIME-LIMIT

Effective date: 20170705

Ref country code: EE

Free format text: LAPSE BECAUSE OF FAILURE TO SUBMIT A TRANSLATION OF THE DESCRIPTION OR TO PAY THE FEE WITHIN THE PRESCRIBED TIME-LIMIT

Effective date: 20170705

Ref country code: SK

Free format text: LAPSE BECAUSE OF FAILURE TO SUBMIT A TRANSLATION OF THE DESCRIPTION OR TO PAY THE FEE WITHIN THE PRESCRIBED TIME-LIMIT

Effective date: 20170705

26N No opposition filed

Effective date: 20180406

PG25 Lapsed in a contracting state [announced via postgrant information from national office to epo]

Ref country code: SI

Free format text: LAPSE BECAUSE OF FAILURE TO SUBMIT A TRANSLATION OF THE DESCRIPTION OR TO PAY THE FEE WITHIN THE PRESCRIBED TIME-LIMIT

Effective date: 20170705

PG25 Lapsed in a contracting state [announced via postgrant information from national office to epo]

Ref country code: MC

Free format text: LAPSE BECAUSE OF FAILURE TO SUBMIT A TRANSLATION OF THE DESCRIPTION OR TO PAY THE FEE WITHIN THE PRESCRIBED TIME-LIMIT

Effective date: 20170705

REG Reference to a national code

Ref country code: BE

Ref legal event code: MM

Effective date: 20180430

PG25 Lapsed in a contracting state [announced via postgrant information from national office to epo]

Ref country code: LU

Free format text: LAPSE BECAUSE OF NON-PAYMENT OF DUE FEES

Effective date: 20180426

PG25 Lapsed in a contracting state [announced via postgrant information from national office to epo]

Ref country code: BE

Free format text: LAPSE BECAUSE OF NON-PAYMENT OF DUE FEES

Effective date: 20180430

PG25 Lapsed in a contracting state [announced via postgrant information from national office to epo]

Ref country code: FR

Free format text: LAPSE BECAUSE OF NON-PAYMENT OF DUE FEES

Effective date: 20180430

PG25 Lapsed in a contracting state [announced via postgrant information from national office to epo]

Ref country code: MT

Free format text: LAPSE BECAUSE OF NON-PAYMENT OF DUE FEES

Effective date: 20180426

PG25 Lapsed in a contracting state [announced via postgrant information from national office to epo]

Ref country code: TR

Free format text: LAPSE BECAUSE OF FAILURE TO SUBMIT A TRANSLATION OF THE DESCRIPTION OR TO PAY THE FEE WITHIN THE PRESCRIBED TIME-LIMIT

Effective date: 20170705

PG25 Lapsed in a contracting state [announced via postgrant information from national office to epo]

Ref country code: PT

Free format text: LAPSE BECAUSE OF FAILURE TO SUBMIT A TRANSLATION OF THE DESCRIPTION OR TO PAY THE FEE WITHIN THE PRESCRIBED TIME-LIMIT

Effective date: 20170705

PG25 Lapsed in a contracting state [announced via postgrant information from national office to epo]

Ref country code: RO

Free format text: LAPSE BECAUSE OF FAILURE TO SUBMIT A TRANSLATION OF THE DESCRIPTION OR TO PAY THE FEE WITHIN THE PRESCRIBED TIME-LIMIT

Effective date: 20170705

Ref country code: MK

Free format text: LAPSE BECAUSE OF NON-PAYMENT OF DUE FEES

Effective date: 20170705

Ref country code: HU

Free format text: LAPSE BECAUSE OF FAILURE TO SUBMIT A TRANSLATION OF THE DESCRIPTION OR TO PAY THE FEE WITHIN THE PRESCRIBED TIME-LIMIT; INVALID AB INITIO

Effective date: 20160426

Ref country code: CY

Free format text: LAPSE BECAUSE OF FAILURE TO SUBMIT A TRANSLATION OF THE DESCRIPTION OR TO PAY THE FEE WITHIN THE PRESCRIBED TIME-LIMIT

Effective date: 20170705

PG25 Lapsed in a contracting state [announced via postgrant information from national office to epo]

Ref country code: AL

Free format text: LAPSE BECAUSE OF FAILURE TO SUBMIT A TRANSLATION OF THE DESCRIPTION OR TO PAY THE FEE WITHIN THE PRESCRIBED TIME-LIMIT

Effective date: 20170705

PGFP Annual fee paid to national office [announced via postgrant information from national office to epo]

Ref country code: DE

Payment date: 20230503

Year of fee payment: 8

Ref country code: CH

Payment date: 20230505

Year of fee payment: 8

PGFP Annual fee paid to national office [announced via postgrant information from national office to epo]

Ref country code: IE

Payment date: 20230720

Year of fee payment: 8

Ref country code: GB

Payment date: 20230502

Year of fee payment: 8