-
公开(公告)号：AU2003278623A1
公开(公告)日：2004-09-30
申请号：AU2003278623
申请日：2003-11-11
Applicant: TANDBERG TELECOM AS
Inventor： MARTON TRYGVE FREDERIK , AARNES INGVAR FLATEN
Abstract: An echo canceller processing echo, noise and near end talk in a narrower, but still intelligible, frequency band for reducing required processing power and complexity. In a preferred embodiment of the present invention, an input audio signal of captured sound in an audio communication system is decimated and then divided into a number of sub bands by an analyze filter. Each sub band is processed as in background audio echo cancelling by subtracting the signal with an echo estimate from a model of the acoustic signal in the respective sub band, except from that the signal is also bypassed, adjusted by a filter and subtracted from the processed signal. The resulting signals are then recombined by a synthesize filter and interpolated to the original sampling rate and bandwidth. Finally, the output from the synthesize filter is added to the input audio signal, which has been delayed and adjusted by a filter. The filters are controlled by a control algorithm detecting the presence of near end sound, far end sound and noise, so that the filters, and consequently the high pass filter of the echo canceller, only pass high frequency (above low pass frequencies) when only near end sound is detected.
-
公开(公告)号：ES2333873T3
公开(公告)日：2010-03-02
申请号：ES04808892
申请日：2004-12-22
Applicant: TANDBERG TELECOM AS
Inventor： MARTON TRYGVE FREDERIK , AARNES INGVAR FLATEN , WINSVOLD BJORN , OVERBY GEIR OLE
IPC: H04M9/08
Abstract: Cancelador de eco de audio ajustado para proporcionar una señal de salida atenuada de eco a partir de una señal de entrada añadida de eco incluye audio de un extremo cercano además de audio de un extremo lejano, añadiendo una contribución de eco, comprendiendo un módulo que genera una primera señal de salida de módulo (5134), configurado por lo menos para implementar un primer modelo adaptativo (5121) de eco acústico, convergiendo para minimizar una parte mono de la contribución de eco en la primera señal de salida de módulo (5134) cuando se proporciona una estimación de eco mono y se resta la estimación de eco mono (5133) de la señal de entrada añadida de eco, creando la primera señal de salida de módulo (5134), y por lo menos un módulo que genera una segunda señal de salida (5136), configurado por lo menos para implementar un segundo modelo adaptativo (5129), que converge para minimizar una parte de audio de múltiples componentes de la contribución de eco en la segunda señal de salida del módulo (5136) cuando se proporciona una estimación de la parte de múltiples componentes (5138) desde la primera señal de salida del módulo (5134), creando la segunda señal de salida del módulo (5136), caracterizado por uno o más detectores de audio de múltiples componentes configurados para detectar si está presente audio de múltiples componentes en el audio del extremo alejado, y en el que el primer modelo adaptativo (5121) está configurado para suspender la convergencia cuando dicho uno con más detector de audio de múltiples componentes detecta substancialmente audio de múltiples componentes en el audio del extremo alejado, y un segundo modelo adaptativo (5129) está configurado para suspender la convergencia cuando dicho uno o más detector de audio de múltiples componentes no detecta substancialmente ningún audio de múltiples componentes en el audio del extremo alejado, o una unidad de procesamiento miscelánea (5122) que monitorizar el audio del extremo alejado y del extremo cercano configurada para atenuar la primera o la segunda señal de salida del módulo (5134, 5136) en eventos de presencia de audio del extremo alejado/extremo cercano de definidos que se activa cuando dicho uno o más detector de audio de múltiples componentes no detecta substancialmente ningún audio de múltiples componentes en el audio del extremo alejado.
-
公开(公告)号：DE602004023968D1
公开(公告)日：2009-12-17
申请号：DE602004023968
申请日：2004-12-22
Applicant: TANDBERG TELECOM AS
Inventor： MARTON TRYGVE FREDERIK , AARNES INGVAR FLATEN , WINSVOLD BJOERN , OVERBY GEIR OLE
IPC: H04M9/08
Abstract: An audio communication system and method with improved acoustic characteristics features a stereo detector that is introduced in the echo canceller of the system. When stereo in far-end audio is detected, converging of the adaptive mono model of the canceller is suspended, and when stereo in far-end audio is not detected, converging of the adaptive stereo model of the canceller is suspended. The system may also be extended with a miscellaneous processing unit configured to attenuate the signal at certain events implying a large stereo echo contribution. A stereo collapsing unit is also introduced on the channels of the far-end audio to remove the stereo image at certain events to further suppress the echo contribution.
-
公开(公告)号：ES2328032T3
公开(公告)日：2009-11-06
申请号：ES03770158
申请日：2003-11-11
Applicant: TANDBERG TELECOM AS
Inventor： MARTON TRYGVE FREDERIK , AARNES INGVAR FLATEN
Abstract: Cancelador de eco de audio para proporcionar una señal de salida de eco atenuado (5155) a partir de una señal de entrada de eco añadido (5152), comprendiendo dicho aparato de cancelación de eco: un primer filtro de análisis (5126) adaptado a dividir una primera señal de entrada de filtro de análisis en un número de primeras sub-señales de entrada de respectivas sub-bandas de frecuencia; y una unidad de procesamiento (5000) adaptada para generar un número de sub-señales de salida basadas en dicho número de primeras sub-señales de entrada mediante la generación, para cada primera sub-señal de entrada (5132), una correspondiente sub-señal de salida (5135); y un filtro de síntesis (5127) adaptado para combinar dicho número de sub-señales de salida en una señal de salida de filtro sintetizador; donde dicha unidad de procesamiento comprende un módulo que incluye, para cada primera subseñal de entrada (5132), correspondientes medios de estimación de eco adaptados para generar una correspondiente sub-señal estimada de eco (5133) mediante la implementación de un modelo de eco acústico y correspondientes medios de sustracción adaptados para generar una correspondiente sub-señal eco-reducida (5134) sustrayendo dicha correspondiente sub-señal estimada de eco (5133) de dicha primera sub-señal de entrada (5132), dicho módulo adaptado además para generar una correspondiente sub-señal de módulo de salida (5150) basada en dicha correspondiente sub-señal eco-reducida (5134), estando dicho cancelador de eco de audio caracterizado por el hecho de que la unidad de procesamiento (5000) comprende además, para cada primera sub-señal de entrada (5132), medios adaptados para generar una correspondiente sub-señal de entrada ajustada mediante la aplicación de un primer filtro (5136) a dicha primera sub-señal de entrada (5132) y medios adaptados para generar dicha correspondiente sub-señal de salida (5135) mediante la sustracción de la correspondiente sub-señal de entrada ajustada de la correspondiente sub-señal de salida de módulo (5150); y por el hecho de que el cancelador de audio comprende además: un primer decimador adaptado para generar la primera señal de entrada de filtro de análisis mediante la aplicación de un filtro de paso bajo (5142) más un sub-muestreador (5141) a la señal de entrada con eco añadido (5152); y un interpolador adaptado a generar una señal de salida de interpolador (5154) mediante la aplicación de un sobremuestreador (5140) más un filtro de paso bajo (5139) a la señal de salida de filtro de síntesis (5139); y un segundo filtro (5138) adaptado para generar una señal de entrada de entrada de eco añadido (5153) filtrando dicha señal de entrada de eco añadido (5152); y un dispositivo de adición adaptado para generar dicha señal de salida de eco atenuado (5155) añadiendo dicha señal de entrada de eco añadido ajustada (5153) a dicha señal de salida de interpolador (5154).
-
公开(公告)号：NO318401B1
公开(公告)日：2005-03-14
申请号：NO20031103
申请日：2003-03-10
Applicant: TANDBERG TELECOM AS
Inventor： MARTON TRYGVE FREDRIK , AARNES INGVAR FLATEN
Abstract: An echo canceller processing echo, noise and near end talk in a narrower, but still intelligible, frequency band for reducing required processing power and complexity. In a preferred embodiment of the present invention, an input audio signal of captured sound in an audio communication system is decimated and then divided into a number of sub bands by an analyze filter. Each sub band is processed as in background audio echo cancelling by subtracting the signal with an echo estimate from a model of the acoustic signal in the respective sub band, except from that the signal is also bypassed, adjusted by a filter and subtracted from the processed signal. The resulting signals are then recombined by a synthesize filter and interpolated to the original sampling rate and bandwidth. Finally, the output from the synthesize filter is added to the input audio signal, which has been delayed and adjusted by a filter. The filters are controlled by a control algorithm detecting the presence of near end sound, far end sound and noise, so that the filters, and consequently the high pass filter of the echo canceller, only pass high frequency (above low pass frequencies) when only near end sound is detected.
-
Patent Agency Ranking
公开(公告)号：NO320942B1
公开(公告)日：2006-02-13
申请号：NO20035796
申请日：2003-12-23
Applicant: TANDBERG TELECOM AS
Inventor： MARTON TRYGVE FREDRIK , AARNES INGVAR FLATEN , WINSVOLD BJORN , OVERBY GEIR OLE
Abstract: An audio communication system and method with improved acoustic characteristics features a stereo detector that is introduced in the echo canceller of the system. When stereo in far-end audio is detected, converging of the adaptive mono model of the canceller is suspended, and when stereo in far-end audio is not detected, converging of the adaptive stereo model of the canceller is suspended. The system may also be extended with a miscellaneous processing unit configured to attenuate the signal at certain events implying a large stereo echo contribution. A stereo collapsing unit is also introduced on the channels of the far-end audio to remove the stereo image at certain events to further suppress the echo contribution.
-
公开(公告)号：DE60327895D1
公开(公告)日：2009-07-16
申请号：DE60327895
申请日：2003-11-11
Applicant: TANDBERG TELECOM AS
Inventor： MARTON TRYGVE FREDERIK , AARNES INGVAR FLATEN
Abstract: An echo canceller processing echo, noise and near end talk in a narrower, but still intelligible, frequency band for reducing required processing power and complexity. In a preferred embodiment of the present invention, an input audio signal of captured sound in an audio communication system is decimated and then divided into a number of sub bands by an analyze filter. Each sub band is processed as in background audio echo cancelling by subtracting the signal with an echo estimate from a model of the acoustic signal in the respective sub band, except from that the signal is also bypassed, adjusted by a filter and subtracted from the processed signal. The resulting signals are then recombined by a synthesize filter and interpolated to the original sampling rate and bandwidth. Finally, the output from the synthesize filter is added to the input audio signal, which has been delayed and adjusted by a filter. The filters are controlled by a control algorithm detecting the presence of near end sound, far end sound and noise, so that the filters, and consequently the high pass filter of the echo canceller, only pass high frequency (above low pass frequencies) when only near end sound is detected.
-
公开(公告)号：NO20035796L
公开(公告)日：2005-06-24
申请号：NO20035796
申请日：2003-12-23
Applicant: TANDBERG TELECOM AS
Inventor： MARTON TRYGVE FREDRIK , AARNES INGVAR FLATEN , WINSVOLD BJORN , OVERBY GEIR OLE
Abstract: An audio communication system and method with improved acoustic characteristics features a stereo detector that is introduced in the echo canceller of the system. When stereo in far-end audio is detected, converging of the adaptive mono model of the canceller is suspended, and when stereo in far-end audio is not detected, converging of the adaptive stereo model of the canceller is suspended. The system may also be extended with a miscellaneous processing unit configured to attenuate the signal at certain events implying a large stereo echo contribution. A stereo collapsing unit is also introduced on the channels of the far-end audio to remove the stereo image at certain events to further suppress the echo contribution.
-
公开(公告)号：NO20035796A
公开(公告)日：2005-06-24
申请号：NO20035796
申请日：2003-12-23
Applicant: TANDBERG TELECOM AS
Inventor： MARTON TRYGVE FREDRIK , AARNES INGVAR FLATEN , WINSVOLD BJORN , OVERBY GEIR OLE
-
公开(公告)号：NO20031103D0
公开(公告)日：2003-03-10
申请号：NO20031103
申请日：2003-03-10
Applicant: TANDBERG TELECOM AS
Inventor： MARTON TRYGVE FREDERIK , AARNES INGVAR FLATEN
Abstract: An echo canceller processing echo, noise and near end talk in a narrower, but still intelligible, frequency band for reducing required processing power and complexity. In a preferred embodiment of the present invention, an input audio signal of captured sound in an audio communication system is decimated and then divided into a number of sub bands by an analyze filter. Each sub band is processed as in background audio echo cancelling by subtracting the signal with an echo estimate from a model of the acoustic signal in the respective sub band, except from that the signal is also bypassed, adjusted by a filter and subtracted from the processed signal. The resulting signals are then recombined by a synthesize filter and interpolated to the original sampling rate and bandwidth. Finally, the output from the synthesize filter is added to the input audio signal, which has been delayed and adjusted by a filter. The filters are controlled by a control algorithm detecting the presence of near end sound, far end sound and noise, so that the filters, and consequently the high pass filter of the echo canceller, only pass high frequency (above low pass frequencies) when only near end sound is detected.
-
-
-
-
-
-
-
-
-
Search Results
10
records
(took 0.019 seconds)
