公开(公告)号：EP1395080A1

公开(公告)日：2004-03-03

申请号：EP02425541.6

申请日：2002-08-30

Applicant: STMicroelectronics S.r.l.

Inventor： Poluzzi, Rinaldo ,  Savi, Alberto ,  Martina, Giuseppe ,  Vago, Davide

IPC: H04R3/00

CPC classification number: H04R3/005 ,  H04R25/407 ,  H04R25/507 ,  H04R2225/41 ,  H04R2410/05

Abstract: A device for filtering electrical signals has a number of inputs (2L, 2R) arranged spatially at a distance from one another and supplying respective pluralities of input signal samples. A number of signal processing channels (10L, 10R), each formed by a neuro-fuzzy filter, receive a respective plurality of input signal samples and generate a respective plurality of reconstructed samples. An adder (11) receives the pluralities of reconstructed samples and adds them up, supplying a plurality of filtered signal samples. In this way, noise components are shorted. When activated by an acoustic scenario change recognition unit (5), a training unit (4) calculates the weights of the neuro-fuzzy filters, optimizing them with respect to the existing noise.

Abstract translation: 一种用于过滤电信号设备具有多个彼此和供给输入信号采样的多个respectivement在距离空间上布置的输入（2L，2R）。 多个信号处理通道（10L，10R），每个由一个神经模糊滤波器构成，接收输入信号采样的respectivement多元性并生成重构样本respectivement多元性。 加法器（11）接收重构样本的复数，并且将它们相加后，供给滤波的信号采样的复数。 通过这种方式，噪声成分被短路。 当由声学场景变化识别单元（5）激活，锻炼单元（4）计算神经模糊滤波器的权重，相对于现有的噪声优化它们。

公开(公告)号：EP1211636A1

公开(公告)日：2002-06-05

申请号：EP00830782.9

申请日：2000-11-29

Applicant: STMicroelectronics S.r.l.

Inventor： Poluzzi, Rinaldo ,  Mione, Cristoforo ,  Savi, Alberto

IPC: G06N3/04

CPC classification number: G06K9/0051 ,  G06N3/0436

Abstract: The filtering device (80) comprises a neuro-fuzzy filter (1; 80) and implements a moving-average filtering technique in which the weights for final reconstruction of the signal ( oL 3 ( i )) are calculated in a neuro-fuzzy network (3) according to specific fuzzy rules. The fuzzy rules operate on three signal features ( X 1( i ), X 2( i ), X 3( i )) for each input sample ( e ( i )). The signal features are correlated to the position of the sample in the considered sample window, to the difference between a sample and the sample at the center of the window, and to the difference between a sample and the average of the samples in the window. The filter device for the analysis of a voice signal comprises a bank of neuro-fuzzy filters (86, 87). The signal is split into a number of sub-bands, according to wavelet theory, using a bank of analysis filters including a pair of FIR QMFs ( H 0 , H 1 ) and a pair of downsamplers (85, 86); each sub-band signal is filtered by a neuro-fuzzy filter (86, 87), and then the various sub-bands are reconstructed by a bank of synthesis filters including a pair of upsamplers (88, 89), a pair of FIR QMFs ( G 0 , G 1 ), and an adder node (92).

Abstract translation: 过滤装置（80）包括一个神经模糊过滤器（1; 80），并实现移动平均滤波技术，其中用于信号的最终的重建（OL3（i））的权重的神经模糊网络计算（ 3）雅丁具体模糊规则。 模糊规则对三个信号特征（X1（i）中，X2（i）中，X3（i））的每一个输入样本（E（i））的操作。 所述信号特征是相关的，以在所考虑的样品窗口中的样本的位置，到样品，并在该窗口的中心的样本之间的差值，以及在样品和样品中的窗口的平均之间的差。 用于语音信号的分析的过滤器装置包括神经模糊滤波器（86，87）的一组。 的信号被分成多个子频带，gemäß的小波理论，使用分析滤波器包括一对FIR QMFs（H0，H1）的一组和一对向下取样器（85，86）; 每个子带信号由神经模糊滤波器（86，87）过滤，并通过合成滤波器包括一对的银行，则各个子带重构上采样器（88，89），一对FIR QMFs的 （G0，G1），以及加法器节点（92）。