Cisco Technology, Inc.

Pexip As

Cisco Systems International S.a.r.l.

Knut Inge Hvidsten

Ragnvald Balch Barth

Michael Anthony Ramalho

Håvard Graff

Tulio Beloqui

Espen Moberg

A method and audio processing device for voice anonymization in an audio- or videoconferencing session. The method comprises receiving a plurality of input audio samples comprising speech, calculating a frequency spectrum of each the plurality of input audio samples, calculating a smoothed spectral magnitude envelope of a first of the plurality of frequency spectrums to determine a plurality of formant features of the speech, each of the plurality of formant features being located at different frequencies in the frequency spectrum, determining one random scaling factor for the audio- or videoconferencing session, determining, based on the one random scaling factor, a voice anonymization function shifting the formant location of at least one of the plurality of formants, and applying the voice anonymization function on the frequency spectrum of each the subsequent plurality of input audio samples in the audio- or videoconferencing session.

Method and audio processing device for voice anonymization

A system and method for determining congestion of a communication link transmitting a media stream over the communication link from a sender device to a receiving device.

Method, system and computer program product for determining congestion of a communication link transmitting a media stream over the communication link

A system and method of downspeeding a media stream transmitted over a communication link from a sender device to a receiving device.

Method, system and computer program product for initiating downspeeding in a videoconferencing session

Doppler correlators are configured to receive samples of a signal sampled based on a frequency. Each Doppler correlator includes successive butterfly elements. Each butterfly element includes cross-coupled first and second branches that include a sample delay that doubles for each successive butterfly element, and a sample inversion selectively placed in one of the first and second branches to encode into the successive butterfly elements of each Doppler correlator the same code sequence. Each Doppler correlator is configured with a respective phase rotation that varies across the Doppler correlators. Each Doppler correlator is configured to correlate the samples against the code sequence and apply the respective phase rotation to the samples as the samples are shifted through the successive butterfly elements, to produce respective correlation results from each Doppler correlator centered on a respective frequency offset from the frequency that varies across the Doppler correlators based on the phase rotations.

Efficient handling of clock offset in spread spectrum decoders

In one example, a first headset establishes a connection with a second headset that is associated with a target participant. The first headset obtains, via the connection, a first audio signal corresponding to speech of the target participant. Based on the first audio signal, one or more parameters associated with a position or a movement of a head of a user of the first headset, one or more head-related transfer functions associated with a shape of the head of the user, and a layout of the environment of the first headset, the first headset modifies the first audio signal to produce a first modified audio signal that corresponds to the speech of the target participant that would be present at the head of the user in absence of the first headset and the noise generated in the environment.

Headset-enabled ad-hoc communication

An apparatus includes an input node, an output node, and successive butterfly elements connected between the input node and the output node. Each butterfly element includes a first branch and a second branch that are cross-coupled with each other and that perform, collectively, sample add and sample delay operations. Either the first branch or the second branch of each butterfly element performs a sample inversion, such that a pattern of the sample inversions across the butterfly elements encodes into the butterfly elements a pattern of negative and positive binary values of a particular row of a Prometheus Orthonormal Sets (PONS) matrix. As a result, the successive butterfly elements correlate input samples applied to the input node against the particular row of the PONS matrix as the input samples are shifted through the successive butterfly elements.

Fast correlation of prometheus orthonormal sets (PONS) for communications

Time-offset, time-overlapping signals are received. The signals each include a pilot code, and at least some of the signals each include a user code occupying a time slot time-synchronized to a respective pilot code. Time-offset cross-correlation peaks for respective ones of the pilot codes are generated, each cross-correlation peak indicating a respective one of the time slots. For each time slot a respective projection vector including user code projections each indicative of whether a respective user code of known user codes is present in the time slot is generated. Particular ones of the projection vectors are selectively combined into an aggregate projection vector of aggregate user code projections, such that the aggregate projection vector has a signal-to-noise ratio (SNR) greater than the projection vectors individually. The user code is selected from among the known user codes based on the aggregate user code projections of the aggregate projection vector.

Adaptive decoding of spread spectrum signals using multiple correlator peaks

A communication device includes a loudspeaker to transmit sound into a room. A signal having a white noise-like frequency spectrum spanning a frequency range of human hearing is generated. Auditory thresholds of human hearing for frequencies spanning the frequency range are stored. Respective levels of background noise in the room at the frequencies are determined. The white noise-like frequency spectrum is spectrally shaped to produce a shaped frequency spectrum having, for each frequency, a respective level that follows either the auditory threshold or the level of background noise at that frequency, whichever is greater. The shaped frequency spectrum is transmitted from the loudspeaker into the room.

Generating spectrally shaped sound signal based on sensitivity of human hearing and background noise level

A method in a decoding process for determining full-resolution chroma pixel information (Cx) corresponding to a spatial fraction of a still-image or a video-frame represented by full-resolution luma pixel information (Y) and decimated chroma pixel information (Cxd) decimated by a decimation process, including: receiving the full-resolution luma pixel information at video or image processing apparatus; decimating, at the video or image processing apparatus, the full-resolution luma pixel information (Y) by said decimation process resulting in a decimated spatial luma fraction (Yd); determining, with the video or image processing apparatus, if the decimated chroma pixel information (Cxd) at least approximately can be expressed by {(Yd+shift1)*scale&#x2212;shift2}; storing, in an electronic memory of the video or image processing apparatus, values of scale, shift1, and shift2 that result in a minimum deviation between {(Yd+shift1)*scale&#x2212;shift2} and Cxd; and calculating, with the video or image processing apparatus, {(Y+shift1)*scale&#x2212;shift2} as a first candidate (Cx) for the full-resolution chroma pixel information (Cx).

Growing community of inventors

Oslo, Norway

Knut Inge Hvidsten