Abstract: Apparatuses, methods, and systems for calibrating of an antenna array that uses low-resolution phase shifters, are disclosed. On method includes generating a codebook of phase-shifter setting selections for each of a plurality of antenna elements of an antenna array including communicating a wireless signal between an external calibration antenna and the antenna array through a beam formed by a reference antenna element of the antenna array and an antenna element of the antenna array being calibrated, measuring a signal power of the communicated wireless signal for each of N settings of a digitally selected phase shifter associated with the antenna element of the antenna array being calibrated, and estimating a virtual signal power of each of M settings of the digitally selected phase shifter based on the signal power measurements of the N settings of the digitally selected phase shifter, wherein M is greater than N.

Abstract: Systems, methods, and devices for increasing inference speed of a trained convolutional neural network (CNN). A first computation speed of first filters having a first filter size in a layer of the CNN is determined, and a second computation speed of second filters having a second filter size in the layer of the CNN is determined. The size of at least one of the first filters is changed to the second filter size if the second computation speed is faster than the first computation speed. In some implementations the CNN is retrained, after changing the size of at least one of the first filters to the second filter size, to generate a retrained CNN. The size of a fewer number of the first filters is changed to the second filter size if a key performance indicator loss of the retrained CNN exceeds a threshold.

Abstract: Apparatuses, methods, and systems for grouping antenna elements to enhance an antenna array response resolution, are disclosed. One method includes selecting a plurality of groups of antenna elements from an antenna array, wherein each group includes a plurality of antenna elements of the antenna array, determining an optimal phase setting for a beam directed from each group to a target device, and characterizing each of the groups including adjusting a phase of an electronic signal passing through each of the antenna elements of the group based on the optimal phase setting of the group, adjusting an amplitude ai of the electronic signal the passing through each of the antenna elements of the group to compensate for a loss of antenna array gain of the antenna array due to grouping of the antenna elements, and storing the adjusted phase and adjusted amplitude for each of the antenna element.

Abstract: Embodiments are configured to provide interactive communication functionality including adaptive video processing functionality that can be used to process aspects of a video signal, but the embodiments are not so limited. In an embodiment, components of a video conferencing system can operate to provide a video signal based in part on the use of adaptive processing features which include scaling and/or other pixel processing features. In one embodiment, components of an interactive video system can operate to adaptively manage and control video payload parameters to adapt to various communication conditions associated with a real-time or near-real time interactive video environment.

Abstract: A device may generate a media stream to be shared with other users by building a media graph, comprising a series of interconnected processing units that perform various processing tasks. However, the time involved in generating the media graph may delay the initialization of the media stream, and adjusting properties of the media stream (such as resolution or codec) may result in an interruption of the media stream while a new media graph is built. Instead, a media graph cache may be provided to cache a set of media graphs, which may be interchangeably selected for rapid initialization and adjusting of media stream properties. The media component (e.g., a videocamera) may also be configured to promote rapid adjustments to some media stream properties, while maintaining other properties (e.g., field of view and white balance) for a smooth transition between media stream property sets.

Abstract: Architecture for enhancing the compression (e.g., luma, chroma) of a video signal and improving the perceptual quality of the video compression schemes. The architecture operates to reshape the normal multimodal energy distribution of the input video signal to a new energy distribution. In the context of luma, the algorithm maps the black and white (or contrast) information of a picture to a new energy distribution. For example, the contrast can be enhanced in the middle range of the luma spectrum, thereby improving the contrast between a light foreground object and a dark background. At the same time, the algorithm reduces the bit-rate requirements at a particular quantization step size. The algorithm can be utilized also in post-processing to improve the quality of decoded video.

Abstract: A device may generate a media stream to be shared with other users by building a media graph, comprising a series of interconnected processing units that perform various processing tasks. However, the time involved in generating the media graph may delay the initialization of the media stream, and adjusting properties of the media stream (such as resolution or codec) may result in an interruption of the media stream while a new media graph is built. Instead, a media graph cache may be provided to cache a set of media graphs, which may be interchangeably selected for rapid initialization and adjusting of media stream properties. The media component (e.g., a videocamera) may also be configured to promote rapid adjustments to some media stream properties, while maintaining other properties (e.g., field of view and white balance) for a smooth transition between media stream property sets.

Abstract: Techniques to perform fast motion estimation are described. An apparatus may comprise a motion estimator operative to receive as input a current frame and a reference frame from a digital video sequence. The motion estimator may generate and output a motion vector. The motion vector may represent a change in position between a current block of the current frame and a matching reference block of the reference frame. The motion estimator may utilize an enhanced block matching technique to perform block matching based on stationary and spatially proximate blocks. Other embodiments are described and claimed.

Abstract: Technologies for recovering from dropped frames in the real-time transmission of video over an IP network are provided. A video streaming module receives a notification from a receiving module that a data packet has been lost. The video streaming module determines, based on the type of video frame conveyed in the lost packet and the timing of the lost packet in relation to the sequence of video frames transmitted to the receiving module, whether or not a replacement video frame should be sent to the receiving module. If the video streaming module determines a replacement video frame is warranted, then the video streaming module instructs a video encoding module to generate a replacement video frame and then transmits the replacement video frame to the receiving module.

Abstract: A device may generate a media stream to be shared with other users by building a media graph, comprising a series of interconnected processing units that perform various processing tasks. However, the time involved in generating the media graph may delay the initialization of the media stream, and adjusting properties of the media stream (such as resolution or codec) may result in an interruption of the media stream while a new media graph is built. Instead, a media graph cache may be provided to cache a set of media graphs, which may be interchangeably selected for rapid initialization and adjusting of media stream properties. The media component (e.g., a video camera) may also be configured to promote rapid adjustments to some media stream properties, while maintaining other properties (e.g., field of view and white balance) for a smooth transition between media stream property sets.

Abstract: A perceptual mechanism for residue selection in a video encoder may be provided. The mechanism may comprise a method, system, or device for receiving video frames comprising pluralities of pixels. For each video frame, a sensitivity threshold may be determined for each pixel of a previous video frame. The pixels of the video frame may compared in turn to the pixels of the previous video frame to determine a residue value. The residue value may be compared to the sensitivity threshold such that when the residue value is less than the sensitivity threshold, the pixel data in the video frame may be zeroed out prior to encoding the video frame for transmission.

Abstract: Embodiments are configured to provide interactive communication functionality including adaptive video processing functionality that can be used to process aspects of a video signal, but the embodiments are not so limited. In an embodiment, components of a video conferencing system can operate to provide a video signal based in part on the use of adaptive processing features which include scaling and/or other pixel processing features. In one embodiment, components of an interactive video system can operate to adaptively manage and control video payload parameters to adapt to various communication conditions associated with a real-time or near-real time interactive video environment.

Abstract: Embodiments are configured to provide interactive communication functionality including adaptive video processing functionality that can be used to process aspects of a video signal, but the embodiments are not so limited. In an embodiment, components of a video conferencing system can operate to provide a video signal based in part on the use of adaptive processing features which include scaling and/or other pixel processing features. In one embodiment, components of an interactive video system can operate to adaptively manage and control video payload parameters to adapt to various communication conditions associated with a real-time or near-real time interactive video environment.

Abstract: Architecture that employs texture sensitive temporal filtering to reuse motion estimation information in a realtime encoder. The temporal filter is applied for classified static areas. The architecture reuses the motion estimation results on motion vectors, cost estimates (e.g., sum of absolute difference (SAD)), and edge awareness texture information to apply the temporal filter on the current picture. Filtering can be applied at the pixel level, block level or macroblock level.

Abstract: A method of sharpening video data may include, for at least some pixels in the video data, determining a motion value for a pixel. The sharpness of the pixel may be changed inversely in relation to the motion value of the pixel. Pixels that have higher velocities may be sharpened less than slower moving or stationary pixels in an image of video data.

Abstract: Architecture that employs texture sensitive temporal filtering to reuse motion estimation information in a realtime encoder. The temporal filter is applied for classified static areas. The architecture reuses the motion estimation results on motion vectors, cost estimates (e.g., sum of absolute difference (SAD)), and edge awareness texture information to apply the temporal filter on the current picture. Filtering can be applied at the pixel level, block level or macroblock level.

Abstract: A method and system are disclosed that detect signal artifacts in one or more event signals. The system and method may be used with a patient monitoring apparatus that adapts to a patient's condition and distinguishes between clinically significant changes in the patient's state verse clinically insignificant changes.

Abstract: A method and system are disclosed that detect signal artifacts in one or more event signals. The system and method may be used to estimate whether a monitored signal includes artifacts based upon a statistical analysis using a transform function.

Abstract: Architecture for accelerating video compression by using the motion vectors produced locally by a camera is disclosed. Video frames are captured by the camera (e.g., a webcam) which also computes a motion vector for the frames. Metadata can also be generated that represent an index of motion quality associated with the motion vector. The motion vector is passed to a video compression engine which selectively uses the motion vector directly or alternatively as a seed for a compression and encoding algorithm. This algorithm produces a compressed video frame representing a motion estimate having a selected motion quality index value. In this way, complexity is reduced in the video compression engine, resulting in faster and more efficient video compression. Alternatively, the webcam sends a compressed video bitstream to reduce throughput on the connection and the receiving computing system processes residual information to derive an estimate of the quality index for each macroblock/kernel.

Abstract: Embodiments are configured to provide video conferencing functionality including using region of interest (ROI) features to provide a video signal, but the embodiments are not so limited. In an embodiment, components of a video conferencing system can operate to provide a video signal using pixel data associated with a ROI. In one embodiment, a video conference device can include a detector that can be used to detect human flesh tone regions in a video scene as part of providing a video stream to one or more conference participants.