Nuance Communications, Inc.

Amazon Technologies, Inc.

Facebook, Inc.

Meta Platforms, Inc.

Sridhar Pilli

Mahesh Godavarti

Jose Lainez

Dushyant Sharma

Patrick A Naylor

Biswaranjan Panigrahi

Daniel Almendro Barreda

Carlo Murgia

Ramya Gopalan

Qian-Yu Tang

Jagadeesh Balam

Subhash Chandra Bose Naripeddy

Vamsi Mynampati

Pratik Mukesh Kamdar

Pejman Dehghani

A communication system detects a state of a connected television. The communication system receives a request to establish a communication session with a local communication device via a network. The communication system outputs a reference audio signal to a locally connected audio output device. The communication system captures an ambient audio signal from a local microphone and compares the reference audio signal and the ambient audio signal to determine if the reference audio signal and the ambient audio signal match by meeting a similarity threshold. initiates an action with respect to the communication session based on the comparison.

Method to acoustically detect a state of an external media device using an identification signal

A communication system establishes a communication session with a communication device via a network and captures, via one or more cameras, a local area within a field of view of the one or more cameras. The local area includes one or more audio sources. The communication system receives, from the communication device, a selection of a respective audio source from the one or more audio source in the local area. The selection indicates that audio originating from the respective audio source is to be prioritized over other audio sources in the local area. The communication system tunes one or more microphones based on a depth of the respective audio source relative to the communication system to optimize reception of audio originating at the respective audio source by the one or more microphones.

Smart audio with user input

The present disclosure is directed towards a method for scheduling data packets in a multi-channel packet processing environment. The method may include receiving one or more data packets associated with an incoming signal and inserting the one or more data packets into a queue. The method may further include monitoring a time delay associated with each of the one or more data packets, wherein the time delay indicates a difference between packet arrival and packet departure times. The method may also include sorting the time delay results based upon an increasing order of time delay and determining a total number of data packets associated with each of a plurality of channels. The method may also include scheduling a data packet for processing based upon, at least in part, at least one of the sorted time delay results and the total number of data packets associated with each channel.

System and method for packet scheduling

A system configured to improve speech quality by performing residual echo suppression (RES). The system may detect when double talk is present in a voice conversation and may use different attenuation parameters based on a frequency of audio data. The system may perform RES on the audio data using a low attenuation value for low frequencies and a high attenuation value for high frequencies and determine that double talk is present when a difference in energy level between the low frequencies and the high frequencies is below a threshold. If double talk is present, the RES may generate output audio data using a low attenuation value for low frequencies and a high attenuation value for high frequencies. If double talk is not present, the RES may generate output audio data using a high attenuation value for low frequencies and a high attenuation value for high frequencies.

Methods for detecting double talk

The present disclosure is directed towards an audio conferencing method. Some embodiments may include receiving, at a first mixing device, a first audio stream from one or more participant conferencing devices. The method may further include generating a top-N voice stream at the first mixing device, wherein the top-N voice stream corresponds with at least one top-N talker and wherein the identification of the at least one top-N talker is based upon, at least in part, an activity ranking. The method may also include receiving the top-N voice stream at a centralized mixing device and generating at least one mixed audio stream at the centralized mixing device.

System and method for audio conferencing

The present disclosure is directed towards an audio conferencing method. Some embodiments may include receiving, at a first mixing device, an audio signal from a first user associated with an audio conference. Embodiments may further include processing the audio signal at the first mixing device to generate a processed audio signal and transmitting the processed audio signal to a second mixing device, wherein the first mixing device and the second mixing device are distributed over a network in a cascaded configuration. Embodiments may also include receiving, at the second mixing device, a third audio signal from a second user associated with the audio conference and processing the third audio signal at the second mixing device to generate a second processed audio signal.

The present disclosure is directed towards a method for discontinuous transmission ('DTX') bandwidth reduction. The method may include receiving, at a processor, a frame identified as speech and determining that the frame was mistakenly identified as speech based upon, at least in part, a voice activity detection algorithm. The method may further include labeling the frame as a silence indicator frame.

System and method to reduce transmission bandwidth via improved discontinuous transmission

The present disclosure is directed towards a method for speech intelligibility. The method may include receiving, at one or more computing devices, a first speech input from a first user and performing voice activity detection upon the first speech input. The method may also include analyzing a spectral tilt associated with the first speech input, wherein analyzing includes computing an impulse response of a linear predictive coding ('LPC') synthesis filter in a linear pulse code modulation ('PCM') domain and wherein the one or more computing devices includes an adaptive high pass filter configured to recalculate one or more linear prediction coefficients.

System and method for speech enhancement on compressed speech

The present disclosure is directed towards a process for estimating the signal to noise ratio of a speech signal. The process may include receiving, at a computing device, a speech signal having a bitstream and a signal-to-noise ratio ('SNR') associated therewith. The process may further include estimating the SNR directly from the bitstream or using a partial decoder that is configured to extract one or more parameters, the parameters including at least one of a fixed codebook gain, an adaptive codebook gain, a pitch lag, and a line spectral frequency ('LSF') coefficient.

Growing community of inventors

Fremont, CA, United States of America

Sridhar Pilli