Abstract: Technologies are provided herein for selecting one or more subimages to be processed from a plurality of images received. A subset of images is identified, from the plurality of images received, that include a face image. The face image is assessed to generate assessment information regarding one or more of a size of the face image, a pose of the face in the face image, sharpness of the face image, and contrast quality of the face image. Each aspect of the assessment information is compared with corresponding criteria and, as a result of determining that the assessment information satisfies the criteria, facial recognition processing is performed on the face image.

Abstract: Embodiments are directed towards video enhancement in accordance with time constraint. An example method includes determining a time constraint for transforming low dynamic range (LDR) video content to high dynamic range (HDR) video content, processing the LDR video content to generate instructions for transforming to HDR video content in accordance with the time constraint, rendering the HDR video content based on executing the generated instructions; and producing metadata including the generated instructions for sharing.

Abstract: Systems and methods are provided for determining a candidate identity of a person in video data based on a biometric motion signature corresponding to motion of a face of the person in the video data. Motion information is obtained corresponding to motion of the face in the video data. The motion information is provided as input to a neural network and the neural network generates the biometric motion signature in response. The biometric motion signature is compared with a plurality of biometric signatures stored in data storage. The candidate identity of the person is determined as a result of a correspondence between the biometric motion signature and a stored biometric signature of the plurality of biometric signatures.

Abstract: Technologies are provided for assessing liveness of a subject presented to a set of one or more cameras. A first image and a second image depicting a face are captured during a first time period, and a third image and a fourth image depicting the face are captured during a second time period. A first variance in facial appearance is detected between the first depiction of the face and the third depiction of the face, and a second variance in facial appearance is detected between the second depiction of the face and the fourth depiction of the face. A liveliness of a person is verified based on a determination that the first variance in facial appearance corresponds in time to the second variance in facial appearance.

Abstract: Embodiments are directed towards hue-based video enhancement. An example method includes processing low dynamic range (LDR) video content to generate an inverse tone map (ITM) for transforming the LDR video content to high dynamic range (HDR) video content, converting the LDR video content into Hue, Saturation and Lightness (HSY) color space to produce H-channel data, S-channel data, and Y-channel data, de-noising the H-channel data, remapping the de-noised H-channel data, the S-channel data, and the Y-channel data based on the ITM; and rendering the HDR video content based thereon.

Abstract: Embodiments are directed towards metadata service for video enhancement. An example method includes receiving a request for metadata applicable to instruct a display device to transform low dynamic range (LDR) video content to high dynamic range (HDR) video content, determining availability of candidate metadata, and responsive to determining that no candidate metadata is available, providing options for generating the metadata applicable to instruct the display device. The example method can also include causing generation of the metadata based at least one of the options, and providing the generated metadata to the display device.

Abstract: Systems and methods are provided for determining a candidate identity of a person in video data based on a biometric motion signature corresponding to motion of a face of the person in the video data. Motion information is obtained corresponding to motion of the face in the video data. The motion information is provided as input to a neural network and the neural network generates the biometric motion signature in response. The biometric motion signature is compared with a plurality of biometric signatures stored in data storage. The candidate identity of the person is determined as a result of a correspondence between the biometric motion signature and a stored biometric signature of the plurality of biometric signatures.

Abstract: Techniques are described herein for processing intra- and inter-messaging platform communications, including by receiving and analyzing messages originating from one sender for distribution to a recipient, where the sender and recipient may be on a same or separate messaging platform. Clusters of such messages with similar contents or other similar characteristics are identified and categorized, such as in accordance with configuration information regarding one or both of the originating and destination messaging platforms. Based on a determination of one or more categories associated with such an identified message cluster, as well as an analysis of metadata associated with the profile of the sender of the messages, various actions may be taken with respect to such message clusters or with parties associated with such message clusters, including actions based at least in part on the configuration information.

Abstract: Systems and methods for enabling a video capture/encoding device to capture frames of a video sequence; obtain feedback from a video decoding/rendering device; divide each frame into an array of patches; select one or more patches of each frame for supplemental processing; cause the frames of the video sequence to be encoded by a video encoding module and cause the selected patches to be processed by one or more supplemental processing module(s); and assemble a multi-layer video transport stream wherein, for each frame of the video sequence, a base layer of the transport stream includes data corresponding to an encoded version of the current frame and one or more supplemental layers of the transport stream include meta-data corresponding to the division of the current frame into patches and data corresponding to the output of the supplemental processing of the selected patches of the current frame.

Abstract: Provided herein are systems and methods for encoding an unencoded video frame of a sequence of video frames using a condensed coding header format. After a frame is divided into coding blocks and the visual data encoded, the coding block header data for each coding block is combined and encoded. A flag is set in the frame's frame header indicating whether the condensed or non-condensed coding block header format is being used.

Abstract: Technologies are provided for assessing liveness of a subject presented to a set of one or more cameras. A pair of face images captured during a first time period are respectively obtained from a first and a second camera. Information regarding each face image of the pair of face images is compared with each other to determine whether the pair of face images are consistent with each other. A third face image captured during a second time period is obtained from the first camera. A first face image of the pair of face images and the third face image are assessed to determine the presence of a particular variance between the first face image and the third face image. A result is provided regarding liveness of the subject presented based on the presence of the particular variance and the determination of consistency between the pair of face images.

Abstract: Provided herein are systems and methods for encoding an array of coefficients representing an unencoded video frame of a sequence of video frames to generate an encoded bit-stream representative of the unencoded video frame. The array of coefficients may have a plurality of coefficient rows and a plurality of coefficient columns and the encoded bit-stream representative of the unencoded video frame may include at least a header and a video data payload. Accordingly, the array of coefficients may be separated into an array of coding blocks; each array of coding blocks may be placed into a coding stripe and each coding stripe may be placed into a coding stripe group. The array of coefficients may then be encoded such that the coefficients of a coding block are encoded using information obtained only from other coefficients in the same coding stripe group and a first video data payload portion of the video data payload corresponding to the coding stripe is an integer number of bytes in length.

Abstract: Technologies are provided herein for selecting one or more subimages to be processed from a plurality of images received. A subset of images is identified, from the plurality of images received, that include a face image. The face image is assessed to generate assessment information regarding one or more of a size of the face image, a pose of the face in the face image, sharpness of the face image, and contrast quality of the face image. Each aspect of the assessment information is compared with corresponding criteria and, as a result of determining that the assessment information satisfies the criteria, facial recognition processing is performed on the face image.

Abstract: A block processing procedure to determine coding-block or transform-block sizes for encoded and decoding video data. An encoder obtains unencoded video data and selects a region-of-interest in the video data. Image characteristics of the video data are determined from the region-of-interest and a block size is determined from those image characteristics. The video data is encoded using the block size without storing the block size in a header associated with the encoded data. A decoder obtains the encoded data and selects a second region-of-interest in the encoded data that corresponds to the unencoded data region-of-interest. Second image characteristics are determined from the second region-of-interest, where the second characteristics are the same as the characteristics determined for the unencoded data. A block size is determined from the second image characteristics without using a header associated with the encoded data. The encoded data is decoded based on the determined block size.

Abstract: Summarization segments of an encoded video can be efficiently identified, without the need to decode the encoded video to obtain image data, by analyzing encoded-buffer-size deltas, each indicating an encoded-buffer-size difference between a pair of intra-coded frames of an encoded video.

Abstract: A transform block processing procedure wherein a maximum coding-block size and a maximum transform-block size for an unencoded video frame is determined. The unencoded video frame is divided into a plurality of coding-blocks including a first coding-block and the first coding block is divided into at least one prediction block and a plurality of transform blocks. The size of the transform blocks depend at least in part on the size of the coding block and the corresponding prediction blocks. The transform blocks are then encoded, thereby generating a video data payload of an encoded bit-stream. A frame header of the encoded bit-stream, including a maximum coding-block size flag and a maximum-transform-block-size flag, is generated.

Abstract: A protocol is provided by which a current block and a neighboring block are identified and the current block is processed. In some variants a deblocking filter is applied with a filtering block size set either to the standard blocksize or to the shared blocksize, depending on whether the shared size of the current and neighboring blocks is smaller than a standard blocksize.

Abstract: An unencoded video frame of a sequence of video frames is encoded to generate an encoded bit-stream representative of the unencoded video frame. The encoded bit-stream includes a header portion and a video data payload portion. The unencoded video frame may be divided into an array of prediction blocks, including a first prediction block. A coding mode is selected from a plurality of coding modes for use in encoding the first prediction block. The first prediction block is encoded using the selected coding mode to generate a portion of the video data payload of the encoded bit-stream. A coding mode selection flag is provided in the header portion of the encoded bit-stream, which indicates which coding mode of the plurality of coding modes was selected for encoding the first prediction block.

Abstract: In a system comprising media files resident on various devices which can interconnect and provide indices of their media library content, an organization method may provide an aggregate view of the media library of all connected devices ordering the media library contents of various connected devices from the point of view of the device from which the view of the aggregate library will be shown. This point-of-access-centric view of the library may be formed by prioritizing one or more of relevance, reliability, or proximity of the connected devices from the perspective of the viewing device.

Abstract: A transform block processing procedure wherein a maximum coding-block size and a maximum transform-block size for an unencoded video frame is determined. The unencoded video frame is divided into a plurality of coding-blocks including a first coding block and the first coding block is divided into at least one prediction block and a plurality of transform blocks. The size of the transform blocks depend at least in part on the size of the coding block and the corresponding prediction blocks. The transform blocks are then encoded, thereby generating a video data payload of an encoded bit-stream. A frame header of the encoded bit-stream, including a maximum coding-block-size flag and a maximum-transform-block-size flag, is generated.