利索-全球专利检索

  1. 登录
  2. 注册

搜索结果

70 条记录 (耗时 0.015 秒)

    SYSTEMS AND METHODS FOR FEATURE EXTRACTION
    4.
    发明申请
    SYSTEMS AND METHODS FOR FEATURE EXTRACTION 审中-公开
    用于特征提取的系统和方法

    公开(公告)号：WO2014210208A1

    公开(公告)日：2014-12-31

    申请号：PCT/US2014/044168

    申请日：2014-06-25

    申请人： QUALCOMM INCORPORATED

    发明人： LU, Wenliang ,  SEN, Dipanjan

    IPC分类号： G10L25/69

    CPC分类号： G10L15/02 ,  G10L25/00 ,  G10L25/69

    摘要： A method for feature extraction by an electronic device is described. The method includes processing speech using a physiological cochlear model. The method also includes analyzing sections of an output of the physiological cochlear model. The method further includes extracting a place-based analysis vector and a time-based analysis vector for each section. The method additionally includes determining one or more features from each analysis vector.

    摘要翻译： 描述了一种通过电子设备进行特征提取的方法。 该方法包括使用生理学耳蜗模型处理语音。 该方法还包括分析生理学耳蜗模型的输出的部分。 该方法还包括为每个部分提取基于位置的分析向量和基于时间的分析向量。 该方法另外包括从每个分析向量确定一个或多个特征。

    HIGH QUALITY DETECTION IN FM STEREO RADIO SIGNALS
    5.
    发明申请
    HIGH QUALITY DETECTION IN FM STEREO RADIO SIGNALS 审中-公开
    FM立体声无线电信号的高质量检测

    公开(公告)号：WO2013045693A2

    公开(公告)日：2013-04-04

    申请号：PCT/EP2012/069330

    申请日：2012-10-01

    申请人： DOLBY INTERNATIONAL AB

    发明人： ENGDEGARD, Jonas ,  PURNHAGEN, Heiko ,  SEHLSTROM, Leif

    IPC分类号： G10L25/69

    CPC分类号： H04H40/45 ,  G10L19/008 ,  G10L25/69 ,  H04B1/1676

    摘要： The present document relates to audio signal processing, in particular to an apparatus and a corresponding method for improving an audio signal of an FM stereo radio receiver. In particular, the present document relates to a method and system for reliably detecting the quality of a received FM stereo radio signal and for selecting an appropriate processing based on the detected quality. An apparatus (20) configured to estimate the quality of a received multi-channel FM radio signal is described. The received multi¬ channel FM radio signal is representable as a mid signal and a side signal, and the side signal is indicative of a difference between a left signal and a right signal. The apparatus (20) comprises a power determination unit configured to determine (101) a power of the mid signal, referred to as mid power, and a power of the side signal, referred to as side power; a ratio determination unit configured to determine (102) a ratio of the mid power and the side power, thereby yielding a mid-to-side ratio; and a quality determination unit configured to determine (105) a quality indicator of the received FM radio signal based at least on the mid-to-side ratio.

    摘要翻译： 本文件涉及音频信号处理，尤其涉及用于改进FM立体声无线电接收器的音频信号的装置和相应方法。 特别地，本文件涉及用于可靠地检测所接收的FM立体声无线电信号的质量并且用于基于检测到的质量选择适当的处理的方法和系统。 描述了被配置为估计所接收的多频道FM无线电信号的质量的装置（20）。 收到的multi＆not; 信道FM无线电信号可表示为中间信号和侧信号，侧信号表示左信号和右信号之间的差异。 所述装置（20）包括功率确定单元，所述功率确定单元被配置为确定（101）被称为中功率的中间信号的功率和被称为侧功率的侧信号的功率; 比率确定单元，被配置为确定（102）中等功率和副功率的比率，从而产生中等比率; 以及质量确定单元，被配置为至少基于所述中侧比确定（105）所接收的FM无线电信号的质量指标。

    METHOD FOR LEARNING AN AUDIO QUALITY METRIC COMBINING LABELED AND UNLABELED DATA
    6.
    发明申请
    METHOD FOR LEARNING AN AUDIO QUALITY METRIC COMBINING LABELED AND UNLABELED DATA 审中-公开

    公开(公告)号：WO2021259842A1

    公开(公告)日：2021-12-30

    申请号：PCT/EP2021/066786

    申请日：2021-06-21

    申请人： DOLBY INTERNATIONAL AB

    发明人： SERRA, Joan ,  PONS PUIG, Jordi ,  PASCUAL, Santiago

    IPC分类号： G10L25/69 ,  G10L25/30

    摘要： Described is a method of training a neural-network-based system for determining an indication of an audio quality of an audio input. The method includes obtaining, as input, at least one training set comprising audio samples. The audio samples include audio samples of a first type and audio samples of a second type, wherein each of the first type of audio samples is labelled with information indicative of a respective predetermined audio quality metric, and wherein each of the second type of audio samples is labelled with information indicative of a respective audio quality metric relative to that of a reference audio sample. The method further includes: inputting the training set to the neural-network-based system; and iteratively training the system to predict the respective label information of the audio samples in the training set.

    通信伝送装置、音声障害検出方法、および、プログラム
    7.
    发明申请
    通信伝送装置、音声障害検出方法、および、プログラム 审中-公开

    公开(公告)号：WO2021171547A1

    公开(公告)日：2021-09-02

    申请号：PCT/JP2020/008269

    申请日：2020-02-28

    申请人： 日本電信電話株式会社

    发明人： 金光　卓生

    IPC分类号： H04M3/22 ,  G10L19/00 ,  G10L21/0208 ,  G10L25/69 ,  H04M1/24

    摘要： 通信伝送装置（１）は、所定時間の音声データ（１００）を、単位時間ごとに区切り、音声レベルが所定の閾値を超えているか否かによりビット列（１０１）に変換する入力音声レベル検出部（１１）と、音声データ（１００）に対して所定の演算処理を行う演算処理部（１０）と、所定時間の演算処理後の音声データ（１００ａ）を、単位時間ごとに区切り、音声レベルが所定の閾値を超えているか否かによりビット列（１０１ａ）に変換する出力音声レベル検出部（１２）と、演算処理前のビット列（１０１）と演算処理後のビット列（１０１ａ）とを比較する所定のロジックに基づき、音声障害が発生しているか否かを判定する比較判定部（１３）とを備える。

    SYSTEM AND METHOD FOR MANAGING VOICE QUERY OF A PRESENTATION
    9.
    发明申请
    SYSTEM AND METHOD FOR MANAGING VOICE QUERY OF A PRESENTATION 审中-公开

    公开(公告)号：WO2021133155A1

    公开(公告)日：2021-07-01

    申请号：PCT/MY2020/050125

    申请日：2020-10-28

    申请人： MIMOS BERHAD

    发明人： SEBASTIAMPI, Chrishanton Vethanayagam A/L ,  CHOONG, Khong Neng ,  KOK, Gin Xian

    IPC分类号： G10L15/08 ,  G10L15/26 ,  G06F16/63 ,  G06F16/61 ,  G11B20/10 ,  G10L25/69

    摘要： The invention relates to a system and method for managing voice query of a presentation. The system includes a client module (101), a moderating module (102), a transcriber module (103), an output display module (104), a controller module (105), a server (106) and a repository (107). The method involves generating a voice query through a client module (201) by recording an audio input; converting the audio input to text format (202); accessing the audio input stored in the server (203); retrieving the converted audio in text format stored in the server (204); and transmitting the converted audio in text format to display the converted audio in text format through at least one computing device (205).