Abstract: This application relates to the field of video processing, and discloses a video stitching method and apparatus, an electronic device, and a non-transitory computer-readable storage medium, where the video stitching method includes: detecting a similarity between a first image and a second image, the first image being an image frame of a first to-be-stitched video, and the second image being an image frame of a second to-be-stitched video; then determining a motion vector of the first image relative to the second image when the similarity meets a preset condition; and then determining at least one compensated frame between the first image and the second image according to the motion vector, and stitching the first image and the second image based on the at least one compensated frame to stitch the first to-be-stitched video and the second to-be-stitched video.

Abstract: Systems, methods, and computer-readable storage media for digital fingerprinting of media content. The system does this by normalizing and pre-processing the media content into a standard format, then generating fingerprints for image, audio, and/or video aspects of the media. These fingerprints are mathematical vectors generated using one or more techniques, such as perceptual hashes coupled with machine learning. Once the fingerprints for new content is generated, the system can compare those fingerprints to fingerprints of known content to determine if the new content contains prohibited content.

Abstract: A method of motion compensation for geometry representation of 3D data is described herein. The method performs motion compensation by first identifying correspondent 3D surfaces in time domain, then followed by a 3D to 2D projection of motion compensated 3D surface patches, and then finally performing 2D motion compensation on the projected 3D surface patches.

Abstract: Systems, methods, and computer-readable storage media for digital fingerprinting of media content. The system does this by normalizing and pre-processing the media content into a standard format, then generating fingerprints for image, audio, and/or video aspects of the media. These fingerprints are mathematical vectors generated using one or more techniques, such as perceptual hashes coupled with machine learning. Once the fingerprints for new content is generated, the system can compare those fingerprints to fingerprints of known content to determine if the new content contains prohibited content.

Abstract: Detecting the start of a credit roll within video program may allow for the automatic extension of video recordings among other functions. The start of the credit roll may be detected by determining the number of text blocks within a sequence of frames and identifying a point in the sequence of frames where a difference between the number of text blocks in frames occurring before the point and the number of text blocks in frames occurring after the point is greatest and exceeds a specified threshold. Text blocks may be identified within each frame by partitioning the frame into one or more segments and recording the segments having a pixel of a sufficiently high contrast. Contiguous segments may be merged or combined into single blocks, which may then be filtered to remove noise and false positives. Additional content may be inserted into the credit roll frames.

Abstract: A sign value of a DC coefficient of a current block is determined using neighbor blocks of the current block. First and second sign values are identified as respectively corresponding to an above neighbor block of the current block and a left neighbor block of the current block. A context value is calculated by applying a first weighting value to the first sign value and a second weighting value to the second sign value. The first weighting value is based on a boundary between the above neighbor block and the current block, and the second weighting value is based on a boundary between the left neighbor block and the current block. A probability value is selected based on the context value. The sign value of the DC coefficient of the current block is then determined using the probability model.

Abstract: A system, device and method of video frame interpolation. In one aspect, a frame interpolation system receive candidate interpolated frame and difference data based on a first frame and second frame. It identifies a high-error region and one or more increasingly lower-error regions of the candidate interpolated frame based on the difference data and applies multiple successive stages of a refinement neural network to the regions of the candidate interpolated frame, from highest- to lowest-error. At each stage, the results of the previous stage are combined with the next lower-error region before being refined by the current stage. In some aspects, the system also generates the candidate interpolated frame and the difference data using a flow estimation neural network.

Abstract: Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for partitioning videos. In one aspect, a method includes obtaining a partition of a video into one or more shots. Features are generated for each shot, including visual features and audio features. The generated features for each shot are provided as input to a partitioning neural network that is configured to process the generated features to generate a partitioning neural network output. The partition of the video into one or more chapters is determined based on the partitioning neural network output, where a chapter is a sequence of consecutive shots that are determined to be taken at one or more locations that are semantically related.

Abstract: A sign value of a DC coefficient of a current block is determined using neighbor blocks of the current block. First and second sign values are identified as respectively corresponding to an above neighbor block of the current block and a left neighbor block of the current block. A context value is calculated by applying a first weighting value to the first sign value and a second weighting value to the second sign value. The first weighting value is based on a boundary between the above neighbor block and the current block, and the second weighting value is based on a boundary between the left neighbor block and the current block. A probability value is selected based on the context value. The sign value of the DC coefficient of the current block is then determined using the probability model.

Abstract: A method of selecting at least one video frame of a video sequence comprising a plurality of video frames. The method determines a time for analysis based on length of the video sequence and processing capability of a running device. A first sampling pattern is based on the determined time for analysis. A first set of frames in the video sequence is sampled using a first sampling pattern. The first set is sampled infrequently throughout the video sequence in accordance with the first sampling pattern. A candidate frame is determined from the sampled frames based on image quality. A second set of the frames comprising one or more of the frames in a narrow range of the video sequence near the determined candidate frame, is determined in accordance with a second sampling pattern. At least one of the video frames is selected from the sampled frames based on image quality.

Abstract: Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for audio classifiers. In one aspect, a method includes obtaining a plurality of video frames from a plurality of videos, wherein each of the plurality of video frames is associated with one or more image labels of a plurality of image labels determined based on image recognition; obtaining a plurality of audio segments corresponding to the plurality of video frames, wherein each audio segment has a specified duration relative to the corresponding video frame; and generating an audio classifier trained using the plurality of audio segment and the associated image labels as input, wherein the audio classifier is trained such that the one or more groups of audio segments are determined to be associated with respective one or more audio labels.

Abstract: This apparatus includes: a video input unit that inputs a first video signal captured at a first frame resolution by a camera, to which a time data item of each frame has been added; a video storage unit that stores the first video data item of each frame of the first video signal in association with the time data item; a position information input unit that receives a position information item of a particular moving object included in the first video signal as a part of a subject; a position information storage unit that stores the received position information item in association with a time; and a controller that reads, from the position information storage unit, the one or more position information items for each particular time interval, calculates a cut region including one or more positions respectively indicated by these information items, cuts the video data item of the cut region from the first video data item of the time interval, and generates the second video data item having the second frame res

Abstract: According to one embodiment, a system is described that comprises a network device adapted to receive a media stream that includes a plurality of packets. The network device of the system processes the media stream, targeted for a client device that is communicatively coupled to the network device, by perform a number of operations. A first operation includes analyzing a packet type for each packet of the plurality of packets and detecting an event causing difficulty for the client device to successfully receive the media stream. Based on both detecting the event and analyzing the packet type for each packet of the plurality of packets, the network device only transmits a portion of the media stream to the client device, where the portion of the media stream is the received media stream absent one or more packets of the plurality of packets.

Abstract: Devices, systems and methods are disclosed for generating video tags associated with previously captured video data. During a recording mode, video data may be captured and uploaded to a server. In response to a command, video may be tagged and selected, for example for inclusion in a video summary. During a standby mode, video data may be captured and temporarily stored in a buffer. In response to the command, video data stored in the buffer prior to the command may be selected. The video data may be selected for further processing and/or uploaded to the server.

Abstract: An electronic device with a display, processor(s), and memory displays a video monitoring user interface including a video feed from a camera located remotely from the client device in a first region and an event timeline in a second region, the event timeline including event indicators for motion events previously detected by the camera. The electronic device detects a user input selecting a portion of the event timeline, where the selected portion of the event timeline includes a subset of the event indicators. In response to the user input, the electronic device causes generation of a time-lapse video clip of the selected portion of the event timeline. The electronic device displays the time-lapse video clip, where motion events corresponding to the subset of the event indicators are played at a slower speed than the remainder of the selected portion of the event timeline.

Abstract: A method and apparatus for an adaptive bit rate system. The method may include streaming media content at a server, determining a discontinuity in the streaming media content based on boundaries corresponding to changes in the streaming media content, and dividing the streaming media content into a plurality of variable duration segments, wherein a duration of one of the plurality of variable duration segments is to be adjusted based on the determined discontinuity to cause the one of the plurality of variable duration segments to end at the determined discontinuity in the streaming media content. The method may further include transmitting the plurality of variable duration segments of the streaming media content from the server.

Abstract: Provided are an apparatus and a method for detecting a scene cut frame. The apparatus and a method for detecting a scene cut frame includes an image selector configured to select input images by n image frames while shifting an input image at h frame intervals; a candidate frame detector configured to detect a candidate frame from the selected n image frames using at least one of a plurality of codecs and a plurality of scene cut detecting algorithms; and a final frame detector configured to detect a frame having a count value higher than a threshold value as a final scene cut frame by performing histogram analysis based on the detected candidate frame, wherein the h and the n are a real number greater than zero and the h is smaller than the n.

Abstract: There is provided a display control device including an operation detection unit configured to detect an enlargement operation on any image displayed within a display screen among a plurality of images having a predetermined relation, and a display control unit configured to determine a display area of the image based on the detected enlargement operation and to cause a portion corresponding to the display area to be enlarged and displayed for each of the images displayed within the display screen.

Abstract: The present invention is a data transfer system that can transfer, with respect to simulation result data, physical quantities necessary for a user, further perform area division based on characteristics of physical quantities, designate an allowable range of a compression error according to a purpose for each of divided areas, compress the data on the basis of the designated error range, transfer the data, and further display an error of each of the divided areas for each of the physical quantities on a receiver/client side.

Abstract: A video recording system including: a camera sensor and a controller configured to: continuously store video in a temporary file storage arrangement from the camera sensor, display a user interface including displaying the video and a record button, upon receiving an activation of the record button at a first time, recording the first time as a start location, updating the user interface to include a stop recording button, upon receiving an activation of the stop recording button, at a second time, marking the second time as a stop location, display a user interface including one or more selectable start time points that precede the first time, receive a selection of a start time point of the start time points, generate a video file from a subset of the temporary file storage arrangement, the video file beginning at a video frame associated with the start time point.

Abstract: A digital video camera comprising: user controls enabling a user to select between at least an up input, a down input, a left input, a right input, and a confirmation input; and a program memory storing instructions to implement a method for trimming a digital video sequence. The method includes: selecting a digital video sequence; initiating a trimming operation; accepting user input to select a start frame and an end frame for a trimmed digital video sequence, wherein the up input and the down input are used to select between a start frame selection mode and an end frame selection mode, and the left input and the right input are used to scroll through the frames of the selected digital video sequence; and trimming the selected video sequence to include the frames between the selected start frame and the selected end frame.

Abstract: The described technology is directed towards user interface technology that guides users in making content (e.g., video selection) and also navigating within selected content. A selection menu view uses style changes to indicate focus changes, e.g., including animating opacity values to indicate a focused item view versus those items not in focus. In another view, a subset of filmstrip views are displayed, each representing a chapter location in a video, along with a flipbook view that represents a selected scene within a selected chapter. Interaction with the filmstrip views changes the visible subset. Interaction with the flipbook view changes the selected scene; a play command plays the video based upon the selected scene.

Abstract: Systems and methods for the manipulation of captured light fields and captured light field image data in accordance with embodiments of the invention are disclosed. In one embodiment of the invention, a system for manipulating captured light field image data includes a processor, a display, a user input device, and a memory, wherein a depth map includes depth information for one or more pixels in the image data, and wherein an image manipulation application configures the processor to display a first synthesized image, receive user input data identifying a region within the first synthesized image, determine boundary data for the identified region using the depth map, receive user input data identifying at least one action, and perform the received action using the boundary data and the captured light field image data.

Abstract: This disclosure describes selection of reference video units to be used for interpolation or extrapolation of a video unit, such as a video frame. A decoder may apply a quality-focused mode to select a reference frame based on quality criteria. The quality criteria may indicate a level of quality likely to be produced by a reference frame. If no reference frames satisfy the quality criteria, interpolation or extrapolation may be disabled. Display of an interpolated or extrapolated frame may be selectively enabled based on a quality analysis. A decoder may apply a resource-focused frame interpolation mode to enable or disable frame interpolation or extrapolation for some frames based on power and quality considerations. In one mode, frame interpolation may be disabled to conserve power when reference frames are not likely to produce satisfactory quality. In another mode, the threshold may be adjustable as a function of power saving requirements.

Abstract: Systems and methods for the manipulation of captured light fields and captured light field image data in accordance with embodiments of the invention are disclosed. In one embodiment of the invention, a system for manipulating captured light field image data includes a processor, a display, a user input device, and a memory, wherein a depth map includes depth information for one or more pixels in the image data, and wherein an image manipulation application configures the processor to display a first synthesized image, receive user input data identifying a region within the first synthesized image, determine boundary data for the identified region using the depth map, receive user input data identifying at least one action, and perform the received action using the boundary data and the captured light field image data.

Abstract: An illustrative example method of processing image data includes dividing a block of data into a plurality of sub-blocks, the block of data corresponding to at least one image; selecting at least one of the sub-blocks according to a predetermined sequence; identifying a plurality of features of interest in the selected sub-block; and processing the selected sub-block to provide information regarding at least one of the features of interest.

Abstract: An information processing apparatus includes an input unit, a storage unit, and a control unit. The input unit receives processed data having resulted from a process in which a test-content is read and processed by a device to be evaluated. The test-content is composed of a plurality of units for reproduction ordered along a time axis, each unit for reproduction having feature points, the number of the feature points monotonically increasing or decreasing by a predetermined difference along passage of reproduction time. The storage unit stores the predetermined difference. The control unit calculates the numbers of the feature points with respect to each of two adjacent units among the units for reproduction in the processed data, and detects an error in the processed data by determining whether or not a difference between the calculated numbers of the feature points is equal to the stored difference.

Abstract: Described are methods and apparatuses, including computer program products, for personalizing video content. A server receives a request for a digital video stream from a client device and generates a first set of personalized frames to be inserted into the requested digital video stream. The server combines the first set of personalized frames and the requested digital video stream to generate a personalized video stream and transmits the personalized video stream to a downstream device. The downstream device generates a second set of personalized frames to be inserted into the personalized video stream and combines the second set of personalized frames and the personalized video stream to generate an updated personalized video stream for transmission to the client device.

Abstract: A system and method for compiling video segments including defining an event; providing a multi-user video aggregation interface; receiving a plurality of video segments through the aggregation interface; determining event-synchronized alignment of the plurality of videos; and assembling a multi-channel video of event, the multi-channel video file configured with at least two video segments that have at least partially overlapping event-synchronized alignment.

Abstract: An input image is partitioned into a plurality of image regions based on color and color differences. The partitioning comprises assigning a color difference value to plurality of locations within the input image. The partitioning further comprises assigning each of the plurality of locations to an image region of the plurality of image regions, where the assigning occurs according to a particular order. The particular order is based at least in part on color difference values associated with the plurality of locations. The input image may comprise markup. Data representing at least a particular portion of the markup in the input image based on the partitioning is identified. Data representing at least the portion of the markup may be used in a visualization of a customizable product or a manufacturing control associated with a customizable product.

Abstract: This disclosure describes techniques for improving functionalities of a back-end device, e.g., a video encoder, using parameters detected by a front-end device, e.g., a video camera. The techniques may involve detecting a scene change in a captured frame, based on one or more parameters of auto exposure (AE), auto white balance (AWB), and auto focus (AF) functions. If a scene change is detected in a captured frame, a video processing device, which may be a stand-alone device, or may be integrated into one of the front-end or back-end devices, provides an indication of the scene change. The video encoder interprets the signal as a trigger to encode the frame indicated as the frame where a scene change occurred as a reference I frame.

Abstract: Quantized energy values are accessed to initially represent a temporally related group of content elements in a media sequence. The values are accessed over a matrix of regions into which the initial representation is partitioned. The initial representation may be downsampled and/or cropped from the content. A basis vector set is estimated in a dimensional space from the values. The initial representation is transformed into a subsequent representation, which is in another dimensional space. The subsequent representation projects the initial representation, based on the basis vectors. The subsequent representation reliably corresponds to the media content portion over a change in a geometric orientation thereof. Repeated for other media content portions of the group, subsequent representations of the first and other portions are averaged or transformed over time. The averaged/transformed values reliably correspond to the content portion over speed changes.

Abstract: A method for generating an interpolation frame between first and second reference frames includes dividing an interpolation frame into several interpolation areas; detecting a most correlated combination from several combinations between first reference areas and second reference areas for each interpolation area; obtaining a motion vector from the first and second reference areas; determining whether the first reference areas and the second reference areas are in a high-correlated area or a low-correlated area; giving the motion vector to the motion vector detected area, the motion vector detected area corresponding to the interpolation area which is determined to be the high-correlated area in the first and second reference areas; determining a motion vector to be given to the motion vector undetected area; and generating the interpolation frame based on the motion vector given to the motion vector detected area and the motion vector determined for the motion vector undetected area.