Abstract: A method includes receiving a plurality of audio signals, wherein the audio signals each include a plurality of packets containing voice data associated with conference participants. One or more of the audio signals is associated with a video signal and at least one of the audio signals is associated with a voice-only conference participant. The method further includes determining an activity metric for each of the plurality of audio signals, wherein the activity metric indicates a confidence value that the audio signal is associated with a conference participant that is currently speaking. The method also includes encoding, for each of the plurality of audio signals, one or more packets of the plurality of packets with the activity metric and sending the plurality of audio signals encoded with an activity metric to a conference bridge.

Abstract: A system for providing location-specific sound in a telepresence system includes a plurality of remote microphones. Each remote microphone is associated with a respective area and operable to generate a sound signal from the voice of at least one user within the respective area. The system also includes a plurality of remote cameras. Each remote camera is associated with a respective remote microphone of the plurality of remote microphones and aligned to generate an image of its associated respective area. The system further includes a plurality of local displays. Each local display is operable to reproduce the image of a respective area generated by a respective remote camera. The system also includes a plurality of local loudspeakers. Each local loudspeaker is positioned proximate to a respective local display and operable to reproduce the sound signal from the voice of the at least one user within the respective area reproduced by the respective local display.

Abstract: A system for providing location-specific sound in a telepresence system includes a plurality of remote microphones. Each remote microphone is associated with a respective area and operable to generate a sound signal from the voice of at least one user within the respective area. The system also includes a plurality of remote cameras. Each remote camera is associated with a respective remote microphone of the plurality of remote microphones and aligned to generate an image of its associated respective area. The system further includes a plurality of local displays. Each local display is operable to reproduce the image of a respective area generated by a respective remote camera. The system also includes a plurality of local loudspeakers. Each local loudspeaker is positioned proximate to a respective local display and operable to reproduce the sound signal from the voice of the at least one user within the respective area reproduced by the respective local display.

Abstract: A method includes receiving a plurality of audio signals, wherein the audio signals each include a plurality of packets containing voice data associated with conference participants. One or more of the audio signals is associated with a video signal and at least one of the audio signals is associated with a voice-only conference participant. The method further includes determining an activity metric for each of the plurality of audio signals, wherein the activity metric indicates a confidence value that the audio signal is associated with a conference participant that is currently speaking. The method also includes encoding, for each of the plurality of audio signals, one or more packets of the plurality of packets with the activity metric and sending the plurality of audio signals encoded with an activity metric to a conference bridge.

Abstract: A system for providing location-specific sound in a telepresence system includes a plurality of remote microphones. Each remote microphone is associated with a respective area and operable to generate a sound signal from the voice of at least one user within the respective area. The system also includes a plurality of remote cameras. Each remote camera is associated with a respective remote microphone of the plurality of remote microphones and aligned to generate an image of its associated respective area. The system further includes a plurality of local displays. Each local display is operable to reproduce the image of a respective area generated by a respective remote camera. The system also includes a plurality of local loudspeakers. Each local loudspeaker is positioned proximate to a respective local display and operable to reproduce the sound signal from the voice of the at least one user within the respective area reproduced by the respective local display.

Abstract: In a speakerphone system, the signal that is directed to a loudspeaker is filtered by a dynamically controllable highpass filter. The filter characteristics are adjusted on the basis of the power of the signal and the degree of recent microphone activity. The filtered signal is a far-talker signal, while the microphone activity is represented as a near-talker signal that includes an echo component as a consequence of loudspeaker-to-microphone coupling. The degree of microphone activity is determined by a subband analyzer. A noise floor estimator may be used to determine the degree to which the near-talker signal is comprised of background noise, so that the noise can be eliminated in the process of controlling the highpass filter.

Abstract: In a speakerphone system, the signal that is directed to a loudspeaker is filtered by a dynamically controllable highpass filter. The filter characteristics are adjusted on the basis of the power of the signal and the degree of recent microphone activity. The filtered signal is a far-talker signal, while the microphone activity is represented as a near-talker signal that includes an echo component as a consequence of loudspeaker-to-microphone coupling. The degree of microphone activity is determined by a subband analyzer. A noise floor estimator may be used to determine the degree to which the near-talker signal is comprised of background noise, so that the noise can be eliminated in the process of controlling the highpass filter.

Abstract: A full duplex speakerphone system and a method of detecting a valid near-end talker component within an outgoing signal utilize a near-talker estimator that produces a near-talker energy estimate using a spectrum analysis. The near-talker energy estimate is indicative of the amount of the valid near-end talker component within the outgoing signal. Depending on the near-talker energy estimate, the system may attenuate the outgoing signal and/or incoming signal and adjust the filter process executed by an adaptive filter of the system in order to provide a clear voice communication between connected parties. In the preferred embodiment, the near-talker estimator operates with at least one processing unit to sample the outgoing signal that has been filtered, or “cancelled,” of echo by an acoustic echo canceller of the system.

Abstract: An echo canceller system and a method of filtering an outgoing signal utilize a volume compensator that operates in tandem with a primary adaptive filter to cancel echo due to a change in echo response of the system within a broadcasting environment. In particular, the volume compensator is designed to quickly adapt to a change in echo response due to an increase or decrease in the broadcast volume of a loudspeaker of the system. The echo canceller system may be incorporated into a personal computer system for a full-duplex speakerphone application. The volume compensator includes a supplementary adaptive filter that operates in parallel with the primary adaptive filter to cancel echo. The primary adaptive filter operates on the outgoing signal to cancel echo that may have been introduced into the outgoing signal along with speech of a near-end caller.

Abstract: A multi-block frequency-domain adaptive filter in an acoustic echo canceller as applied to a full-duplex speakerphone includes convolving a received reference signal with a linear room model to generate an echo free signal for transmission which, in turn, is used to adapt the coefficients of the room model. The frequency-domain estimated echo response is transformed into the time-domain to generate an approximated time-domain echo signal which is subtracted from a microphone signal containing a desired intelligence signal and the actual echo to produce an outgoing signal. The outgoing signal serves as an error signal used to adapt the filter. In the preferred embodiment, a weighting means adjusts the frequency-domain room coefficients during the convolution operation. In another embodiment, a weighting means adjusts the frequency-domain room coefficients by adjusting the correlation terms during adaptation of the frequency-domain room coefficients.