Cognitive Systems Corp.

Ssimwave Inc.

Imax Corporation

Christopher Vytautas Olekas

Dustin Griesdorf

Yunfeng Piao

Mohammad Omer

Stephen Arnold Devison

Tajinder Manku

Oleksiy Kravets

Ahmed Badr

Ian Wormsbecker

Peter Olijnyk

Zhou Wang

Abdul Rehman

Waqas Ahmed

Kai Zeng

Jiheng Wang

Mikhail Alexand Zakharov

Joshua Vijay Mohan

Candidate encodes of a source video are produced using a plurality of encoding configurations. Quality of experience (QoE) scores for each of the candidate encodes are computed. The QoE scores of the plurality of candidate encodes are compared to determine a lowest bitrate encode that achieve a target QoE score. The lowest bitrate encode that achieve the target QoE score as an optimized output video is selected.

Perceptual quality assessment metric based content and viewer aware encoding

Real-time latency of audio/video streams is identified. Signatures of a reference audio/video stream and signatures of a test audio/video stream are buffered. A needle is constructed as a vector including a set of signatures of the reference audio/video stream. Correlations of the needle to successive vectors of sets of signatures of the test audio/video stream are computed using a correlation function that calculates relatedness of the needle vector to each of the successive vectors of the test audio/video stream. A synchronization offset is identified between the test stream and the reference stream according to a maximum correlation point of the correlations of the needle to the successive sets of signatures of the test audio/video stream. The reference audio/video stream and the test audio/video stream are aligned according to the synchronization offset.

Distributed measurement of latency and synchronization delay between audio/video streams

A source quality of a source video and a source content complexity of the source video are identified. Parameter constraints with respect to parameters of an operation are received. The source video quality, source content complexity, and parameter constraints are applied to a deep neural network (DNN) producing DNN outputs. In an example, the DNN outputs are combined using domain knowledge to provide the filter parameters, as predicted, to a filter chain, such that applying the filter chain to the input source video results in an output video achieving the full reference video quality score. In another example, the DNN outputs are combined using domain knowledge to provide the filter parameters, as predicted, to a filter chain, such that applying the filter chain to the input source video results in an output video achieving the full reference video quality score.

Relationship modeling of encode quality and encode parameters based on source attributes

In a general aspect, a location of detected motion in a space is determined. In some aspects, motion of an object in a space is detected based on wireless signals communicated through the space by a wireless communication system that includes multiple wireless communication devices. Each wireless signal is transmitted and received by a respective pair of the wireless communication devices. Motion indicator values are computed for the respective wireless communication devices. The motion indicator value for each individual wireless communication device represents a degree of motion detected by the individual wireless communication device based on a subset of the wireless signals transmitted or received by the individual wireless communication device. A location of the detected motion in the space is determined based on the motion indicator values.

Motion localization in a wireless mesh network based on motion indicator values

In a general aspect, motion is detected based on statistical parameters of received wireless signals. In some aspects, signals are obtained. The signals are based on wireless signals transmitted through a space and received at a wireless communication device. Values of a set of statistical parameters are computed for each signal, and filtered values are obtained by applying a filter to the values of the statistical parameters. The filter includes parameters based on a signal quality analysis of the wireless signals. A motion detection process is executed to determine, based on the filtered values, whether an object moved in the space.

Motion detection based on filtered statistical parameters of wireless signals

In a general aspect, motion in a space can be detected based on machine learning of wireless signal properties. In some aspects, sets of tagged neural network input data are obtained at a neural network training system. Each set of tagged neural network input data is based on a statistical analysis of a series of wireless signals transmitted through a space over a respective time period, and each set of the tagged neural network input data includes a tag indicating whether motion occurred in the space over the respective time period. The sets of tagged neural network input data are processed by the neural network training system to parameterize nodes of a neural network system. Parameterizing the nodes configures the neural network system to detect motion based on untagged neural network input data.

Motion detection based on machine learning of wireless signal properties

In a general aspect, a wireless communication channel is selected based on signal quality metrics. In some aspects, a first wireless communication device receives a first set of signals based on wireless signals transmitted through a space on a set of wireless communication channels from a second wireless communication device. A signal quality metric value is computed for each wireless communication channel based on the first set of signals, and one of the wireless communication channels is selected based on a comparison of the signal quality metric values for the respective wireless communication channels. The first wireless communication device receives a second set of signals based on wireless signals transmitted through the space on the selected wireless communication channel from the second wireless communication device, and a motion detection process is executed to detect motion of an object in the space based on the second set of signals.

Selecting wireless communication channels based on signal quality metrics

In a general aspect, motion is detected using vector representations of channel responses. In some aspects, a first set of channel responses are obtained based on wireless signals transmitted through a space during a first time period. From the first set of channel responses, a set of orthogonal axes in a frequency vector domain are determined. A second channel response is obtained based on a wireless signal transmitted through the space during a second time period, and a channel vector representing the second channel response in the frequency vector domain is determined. Motion of an object in the space is detected based on a projection of the channel vector onto one of the set of orthogonal axes.

Detecting motion based on decompositions of channel response variations

In a general aspect, motion is detected based on statistical parameters of received wireless signals. In some aspects, signals are obtained. The signals are based on wireless signals transmitted through a space and received at a wireless communication device. Values of a set of statistical parameters are computed for each signal, and groupings of the statistical parameter values are identified. A motion detection process that uses the identified groupings is executed to determine whether an object moved in the space.

Motion detection based on groupings of statistical parameters of wireless signals

In a general aspect, a wireless communication channel is selected based on signal quality metrics. In some aspects, a first wireless communication device receives a first set of signals based on wireless signals transmitted through a space on a first wireless communication channel from a second wireless communication device. A signal quality metric value is computed based on the first set of signals, and a second wireless communication channel is selected based on a determination that the value of the signal quality metric does not meet a quality criterion for a motion detection process. The first wireless communication device receives a second set of signals based on wireless signals transmitted through the space on the second wireless communication channel from the second wireless communication device, and the motion detection process is executed to detect motion of an object in the space based on the second set of signals.

In a general aspect, the radio state of a wireless communication device is controlled. In some aspects, a first set of motion detection signals is processed by operation of a radio subsystem of a wireless communication device in a first radio state. An amount of variation in the first set of motion detection signals is determined based on values of a parameter of the motion detection signals. A counter is updated in response to a determination that the variation is greater than a first threshold, and the radio subsystem is changed to a second radio state based on comparing the value of the counter with a second threshold. A motion detection process is executed to detect motion of an object in the space based on a second set of motion detection signals processed by operation of the radio subsystem in the second radio state.

Controlling radio states for motion detection

In a general aspect, a filter time aligns time varying RF-motion data. In some aspects, a method includes loading a first set of input signals (SISs) into an interference buffer over a first time period. The first SISs is based on a first set of wireless signals received by a wireless communication device (WCD) after being transmitted through a space. The method includes using the first SISs to determine an interference level. The method includes: in response to determining the interference level is below a threshold, executing a motion detection process in a second, subsequent time period over which a second SISs is loaded into the interference buffer. The motion detection process, using a subset of the first SISs, detects motion of an object in the space. The second SISs are based on a second set of wireless signals received by the WCD after being transmitted through the space.

Time-alignment of motion detection signals using buffers

In a general aspect, motion detected using wireless signals is categorized. In some aspects, frequency response signals are obtained. The frequency response signals are based on wireless signals that were transmitted through a space and received at a wireless sensor device over a time period. Values of a statistical parameter are determined for the time period, with the statistical parameter for the time period being based on a function applied to frequency components of the frequency response signals over the time period. A category of motion that occurred in the space during the time period is identified based on the values of the statistical parameter.

Growing community of inventors

Breslau, Canada

Christopher Vytautas Olekas