-
21.发明申请
公开(公告)号：WO2021071577A1
公开(公告)日：2021-04-15
申请号：PCT/US2020/045670
申请日：2020-08-10
申请人： BOOMCLOUD 360, INC.
发明人： SELDESS, Zachary
摘要： A system processes an audio signal using spectrally orthogonal sound components. The system includes a circuitry that generates a mid component and a side component from a left channel and a right channel of the audio signal. The circuitry generates a hyper mid component including spectral energy of the side component removed from spectral energy of the mid component, and generate a residual mid component including spectral energy of the hyper mid component removed from the spectral energy of the mid component. The circuitry filters subbands of the residual mid component, such as to apply a subband spatial processing. The circuitry generates a left output channel and a right output channel using the filtered subbands of the residual mid component.
-
公开(公告)号：WO2021060041A1
公开(公告)日：2021-04-01
申请号：PCT/JP2020/034646
申请日：2020-09-14
申请人： ヤマハ株式会社
摘要： 音響信号解析システムは、音響信号の周波数スペクトルを時間軸上で平均した第１スペクトルを取得する取得部と、所定の音律の音高に対応する複数の基準値に対してそれぞれ周波数差を有する複数の成分を含む第２スペクトルであって、前記第１スペクトルとの類似度が所定の閾値を上回る第２スペクトルに対応する周波数差を分割探索により特定する特定部と、特定部により特定された周波数差に含まれる系統誤差が低減されるように当該周波数差を補正する補正部とを具備する。
-
公开(公告)号：WO2020138624A1
公开(公告)日：2020-07-02
申请号：PCT/KR2019/009876
申请日：2019-08-07
申请人： LG ELECTRONICS INC.
发明人： LEE, Bongki , KIM, Youngman , LEE, Keunsang , JEONG, Jaewoong , HWANG, Seunghyun
IPC分类号： G10L21/0208 , G10L21/0272 , G10L25/30 , G10L25/18 , G06N3/08
摘要： An embodiment of the present invention provides an apparatus for noise canceling that includes: an input unit configured to receive an input voice signal that is a target of noise canceling; and a processor configured to generate a first voice signal by canceling noise from the input voice signal on the basis of a noise canceling model which is trained using a plurality of reference voice signals through a deep learning algorithm, generate a second voice signal by canceling residual noise from the first voice signal on the basis of statistical analysis, and generate an output voice signal corresponding to the second voice signal.
-
公开(公告)号：WO2020051451A1
公开(公告)日：2020-03-12
申请号：PCT/US2019/049953
申请日：2019-09-06
申请人： GRACENOTE, INC.
发明人： COOVER, Robert , RAFII, Zafar
IPC分类号： G10L19/018 , G10L19/025 , G10L25/18
摘要： Methods, apparatus, systems, and articles of manufacture are disclosed to fingerprint audio via mean normalization. An example apparatus for audio fingerprinting includes a frequency range separator to transform an audio signal into a frequency domain, the transformed audio signal including a plurality of time-frequency bins including a first time-frequency bin, an audio characteristic determiner to determine a first characteristic of a first group of time-frequency bins of the plurality of time-frequency bins, the first group of time-frequency bins surrounding the first time-frequency bin and a signal normalizer to normalize the audio signal to thereby generate normalized energy values, the normalizing of the audio signal including normalizing the first time-frequency bin by the first characteristic. The example apparatus further includes a point selector to select one of the normalized energy values and a fingerprint generator to generate a fingerprint of the audio signal using the selected one of the normalized energy values.
-
公开(公告)号：WO2019050646A1
公开(公告)日：2019-03-14
申请号：PCT/US2018/045405
申请日：2018-08-06
申请人： CIRRUS LOGIC INTERNATIONAL SEMICONDUCTOR LTD. , HATAB, Ziad Ramez , LINDEMANN, Eric , MELANSON, John, L.
IPC分类号： G10L21/0316 , G10L25/18
CPC分类号： H03G1/04 , G10L21/0316 , G10L25/18 , H03G3/32 , H04R3/00 , H04R3/002 , H04S2400/09 , H04S2400/13
摘要： Audio distortion by a speaker may be reduced by detecting onset audio events within an audio signal and modifying the audio to reduce the audio distortion perceived by a listener. The onsets may be detected using a psych-acoustic model by determining critical sub-band powers and corresponding masking thresholds. When a loudness value calculated from the CSBs and masking thresholds exceeds a threshold level, certain frequency bands may be attenuated and other frequency bands may be amplified. The audio modification may be performed on a frame-by-frame basis and each frame may be processed multiple times until the onset is sufficiently masked or attenuated.
-
公开(公告)号：WO2018157417A1
公开(公告)日：2018-09-07
申请号：PCT/CN2017/076700
申请日：2017-03-15
申请人： GOERTEK. INC
摘要： A signal processing device, a signal processing method, a speaker and an electronics apparatus. The signal processing device comprises a multi-band dynamic range controller, wherein the multi-band dynamic range controller receives an audio signal (S1100) and includes a first band splitting unit and a resonant band adjustment unit; the first band splitting unit is configured to split the audio signal into multiple bands and obtain at least one resonant band therefrom (S1200), which has a resonant frequency band signal in a resonant frequency range of the audio signal; and the resonant band adjustment unit is configured to adjust the resonant frequency band signal based at least on a resonant band dynamic range control gain (S1300) and output an adjusted resonant frequency band signal (S1400) for combination with other band signals into a compression output signal.
-
27.发明申请
公开(公告)号：WO2017195194A1
公开(公告)日：2017-11-16
申请号：PCT/IL2017/050508
申请日：2017-05-09
摘要： The present invention relates to a method for measuring behavioral change in human consciousness that is based on 12 different personality consciousness codes, wherein each code enables to instantly change the state-of-mind of an individual person. The method comprises: a) storing reference voice characteristics of different persons that represent acoustic information as expressed by human voice in a form of a time to frequency component relation; b) classifying the acoustic information into 12 different personality consciousness codes by using support vector machine that analyzes said acoustic information; c) receiving data indicative of a sound energy generated by the voice of said individual; d) performing spectral analysis of said received sound energy in order to obtain voice characteristics from an electronic representation of said sound energy; and e) comparing said obtained voice characteristics with the reference voice characteristics and determining the personality consciousness code of said individual by using the support vector machines, and using the obtained voice characteristics to determine the level of consciousness.
摘要翻译： 本发明涉及一种用于测量基于12种不同人格意识代码的人类意识中的行为变化的方法，其中每种代码能够立即改变个人的精神状态。 该方法包括：a）以时间 - 频率分量关系的形式存储表示由人声表达的声学信息的不同人的参考声音特性; b）通过使用分析所述声学信息的支持向量机将声学信息分类成12种不同的人格意识代码; c）接收指示由所述个人的声音产生的声能的数据; d）对所述接收到的声能进行频谱分析，以便从所述声能的电子表示中获得话音特性; 和e）将所述获得的声音特性与参考声音特性进行比较，并通过使用支持向量机确定所述个人的人格意识代码，并且使用所获得的声音特性来确定意识水平。
-
公开(公告)号：WO2017050175A1
公开(公告)日：2017-03-30
申请号：PCT/CN2016/099053
申请日：2016-09-14
申请人： 阿里巴巴集团控股有限公司 , 杜志军
发明人： 杜志军
摘要： 一种音频识别方法及系统，所述方法包括：对待识别音频文件的语谱图中的第一特征点进行扩散处理，得到特征点图，所述第一特征点的数量为多个（S110）；在目标音频文件的语谱图中查找是否存在与所述特征点图中扩散处理后的各第一特征点分别对应的第二特征点（S120）；若是，则确定所述待识别音频文件为所述目标音频文件的一部分（S130），本方法可以在音频识别中提高特征点的匹配成功率。
-
公开(公告)号：WO2016109491A1
公开(公告)日：2016-07-07
申请号：PCT/US2015/067800
申请日：2015-12-29
发明人： ROT, Moshe , ROTHSCHILD, Lilach , LERNER, Smadar
IPC分类号： G10L25/18
CPC分类号： A61B5/4803 , A61B5/0022 , A61B5/486 , A61B5/7282 , A61B5/742 , A61B5/7465 , G06F19/00 , G06F19/3418 , G06F19/3481 , G09B5/02 , G09B7/00 , G09B19/04 , G10L25/66 , H04L65/00 , H04L67/10
摘要： A method and device for detecting errors when practicing fluency shaping exercises, are presented. The method includes receiving a set of initial energy levels; setting a set of thresholds to their respective initial values; receiving a voice production of a user practicing a fluency shaping exercise; analyzing the received voice production to compute a set of energy levels composing the voice production; detecting based on the computed set of energy levels, the set of initial energy levels, and the set of a threshold of at least one speech-related error, wherein the detection of the at least one speech-related error is respective of the fluency shaping exercise being practiced by the user; and upon detection of the at least one speech-related error, generating a feedback indicating the at least one detected speech-related error.
摘要翻译： 提出了一种在练习流畅形状练习时检测误差的方法和装置。 该方法包括接收一组初始能级; 将一组阈值设置为其各自的初始值; 接受练习流畅形状运动的用户的声音制作; 分析接收到的语音产生，以计算组成语音产生的一组能级; 基于所计算的能量集合，所述初始能级集合以及至少一个语音相关误差的阈值集合进行检测，其中所述至少一个语音相关误差的检测分别对应于所述流畅整形 用户练习; 并且在检测到所述至少一个语音相关错误时，产生指示所述至少一个检测到的语音相关错误的反馈。
-
公开(公告)号：WO2014208319A1
公开(公告)日：2014-12-31
申请号：PCT/JP2014/065197
申请日：2014-06-09
申请人： クラリオン株式会社
发明人： 福江 一智
IPC分类号： G06F17/10 , G10K15/00 , G10L21/0364 , G10L25/18 , H03G5/02
CPC分类号： H03G5/025 , G06F17/14 , H03G5/02 , H03H17/0213
摘要： フィルタ係数群演算装置を、入力手段により入力された周波数特性を逆フーリエ変換する手段と、逆フーリエ変換により得られた数値列を短時間フーリエ変換する手段と、短時間フーリエ変換により得られた周波数領域の信号に対して周波数が高いほど窓長が短くなる関数を用いて窓掛けを行う手段と、窓掛け後の周波数領域の信号を短時間逆フーリエ変換する手段と、短時間逆フーリエ変換により得られた数値列をオーバラップ加算する手段と、オーバラップ加算後の数値列を、入力手段により入力された周波数特性のフィルタをなすフィルタ係数群として決定する手段と、から構成する。
摘要翻译： 滤波器系数组计算装置由以下部分构成：用于对通过输入装置输入的频率特性进行逆傅立叶变换的装置; 用于对通过逆傅里叶变换获得的数字序列应用短期傅里叶变换的装置; 用于应用窗口函数的装置，其中频率越高，窗口长度越短，通过短期傅里叶变换获得的频域的信号; 用于在应用窗口函数之后对频域的信号应用反向短期傅里叶变换的装置; 用于重叠加上通过短期傅立叶变换获得的数字序列的装置; 以及用于确定重叠相加之后的数字序列作为形成由输入装置输入的频率特性的滤波器的滤波器系数组的装置。
-
-
-
-
-
-
-
-
-
搜索结果
208 条记录 (耗时 0.02 秒)
