Abstract: Methods and systems for detection and removal of cast shadows from an image. In particular, one or more embodiments compute correspondences between image patches in the image using a grid-based patch-matching algorithm. One or more embodiments then train a regression model to detect shadows from the computed patch correspondences. One or more embodiments then segment the detected shadows into shadow regions and identify cast shadows from the shadow regions. Once the cast shadows are identified, one or more embodiments use patch-based synthesis of pixels guided by a direct inversion of the image. Optionally, one or more methods can use pixels from the synthesized image and the naïve inversion of the image, based on a synthesis confidence of each pixel, to produce a combined result.

Abstract: Provided herein are compounds that bind to dCK and methods for treating cancer.

Abstract: Techniques are disclosed for removing haze from an image or video by constraining the medium transmission used in a haze image formation model. In particular, a de-hazed scene, which is a function of a medium transmission, is constrained to be greater than or equal to a fractionally scaled variant of the input image. The degree to which the input image is scaled can be selected manually or by using machine learning techniques on a pixel-by-pixel basis to achieve visually pleasing results. Next, the constrained medium transmission is filtered to be locally smooth with sharp discontinuities along image edge boundaries to preserve scene depth. This filtering results in a prior probability distribution that can be used for haze removal in an image or video frame. The input image is converted to gamma decoded sRGB linear space prior to haze removal, and gamma encoded into sRGB space after haze removal.

Abstract: A method for serving a request for hierarchical data includes receiving, at a processing circuit, a request for hierarchical data stored in one or more databases. The request includes an indication of a particular data hierarchy. The method further includes identifying, by the processing circuit, one or more entity count records corresponding to the particular data hierarchy. The entity count records indicate a distribution of the hierarchical data within the data hierarchy. The method further includes splitting the request for hierarchical data into a plurality of sharded requests based on the distribution of the hierarchical data indicated by the one or more entity count records, assigning the plurality of sharded requests to one or more processing sites for processing at least a subset of the sharded requests in parallel, and reporting a combined result of processing the plurality of sharded requests.

Abstract: A method for processing deferred messages includes a storing process, a notification process and a retrieving process. In the storing process, when a user presets and requires to store the deferred message, the system stores the deferred message to be sent to the user; in the notification process, when a user gets online, an IM server at the client side detects the system has stored the deferred message for the user and if the user has subscribed the deferred message notification, the system sends the deferred message notification to the client according to the setting of the system or of the user; in the retrieving process, when the client receives the deferred message notification and requests to acquire the deferred message, its IM server sends the deferred message content to the client. The present invention improves the method for processing deferred messages in the SIP/SIMPLE IM service system.

Abstract: A blurred image having a spatially invariant motion blur resulting from camera motion during image capture is deblurred based on one or more light streaks identified and extracted from the blurred image. A blur kernel for the blurred image is estimated by performing an optimization procedure having a blur kernel constraint based at least in part on the light streak. One or more light streaks can in some embodiments be posed as the blur kernel constraint. A modeled light streak may be defined as a convolution between the blur kernel and a simulated light source, with the optimization procedure being to minimize a distance between the modeled light streak and the corresponding identified light streak from the blurred image.

Abstract: A computer-implemented method and apparatus are described for deblurring an image. The method may include accessing the image that has at least one blurred region and, automatically, without user input, determining a first value for a first size for a blur kernel for the at least one blurred region. Thereafter, automatically, without user input, a second value for a second size for the blur kernel is determined for the at least one blurred region. A suggested size for the blur kernel is then determined based on the first value and the second value.

Abstract: Techniques for detecting and recognizing text may be provided. For example, an image may be analyzed to detect and recognize text therein. The analysis may involve detecting text components in the image. For example, multiple color spaces and multiple-stage filtering may be applied to detect the text components. Further, the analysis may involve extracting text lines based on the text components. For example, global information about the text components can be analyzed to generate best-fitting text lines. The analysis may also involve pruning and splitting the text lines to generate bounding boxes around groups of text components. Text recognition may be applied to the bounding boxes to recognize text therein.

Abstract: Systems and methods are provided for providing improved de-noising image content by using directional noise filters to accurately estimate a blur kernel from a noisy blurry image. In one embodiment, an image manipulation application applies multiple directional noise filters to an input image to generate multiple filtered images. Each of the directional noise filters has a different orientation with respect to the input image. The image manipulation application determines multiple two-dimensional blur kernels from the respective filtered images. The image manipulation application generates a two- two-dimensional blur kernel for the input image from the two-dimensional blur kernels for the filtered images. The image manipulation application generates a de-blurred version of the input image by executing a de-blurring algorithm based on the two-dimensional blur kernel for the input image.

Abstract: A computer-implemented method and apparatus are described for automatically selecting a region in a blurred image for blur kernel estimation. The method may include accessing a blurred image and defining a size for each of a plurality of regions in the image. Thereafter, metrics for at least two of the plurality of regions are determined, wherein the metrics are based on a number of edge orientations within each region. A region is selected from the plurality of regions based on the determined metrics, and a blur kernel for deblurring the blurred image is then estimated for the selected region. The blurred image is then deblurred using the blur kernel.

Abstract: Methods for enhancing images with increased efficiency include using a discriminative index tree to expedite image optimization processes. The discriminative index tree indexes patch-based image priors for modifying an image by using classifiers determined by exploiting a structure of the patch-based image priors. The discriminative index tree quickly and efficiently parses a space of patch-based image patches to determine approximate dominant patch-based image priors for the space of image patches. To further improve the efficiency of the discriminative index tree, one or more embodiments can limit a number of potential patch-based image priors from which a dominant patch-based image prior is selected.

Abstract: Multi-feature image haze removal is described. In one or more implementations, feature maps are extracted from a hazy image of a scene. The feature maps convey information about visual characteristics of the scene captured in the hazy image. Based on the feature maps, portions of light that are not scattered by the atmosphere and are captured to produce the hazy image are computed. Additionally, airlight of the hazy image is ascertained based on at least one of the feature maps. The calculated airlight represents constant light of the scene. Using the computed portions of light and the ascertained airlight, a dehazed image is generated from the hazy image.

Abstract: A computer-implemented method and apparatus are described for deblurring an image. The method may include accessing the image that has at least one blurred region and, automatically, without user input, determining a first value for a first size for a blur kernel for the at least one blurred region. Thereafter, automatically, without user input, a second value for a second size for the blur kernel is determined for the at least one blurred region. A suggested size for the blur kernel is then determined based on the first value and the second value.

Abstract: Techniques are disclosed for collaborative and synchronized photography across multiple digital camera devices. A panoramic photograph of a scene can be generated from separate photographs taken by each of the cameras simultaneously. During composition, the viewfinder images from each camera are collected and stitched together on the fly to create a panoramic preview image. The panoramic preview is then displayed on the camera devices as live visual guidance, which each user can use to change the orientation of the camera and thus change the composition of the panoramic photograph. In some cases, the host sends visual instructions to other camera devices to guide users in camera adjustment. When the desired composition is achieved, the host sends a trigger command to all of the cameras to take photographs simultaneously. Each of these separate photographs can then be stitched together to form a panoramic photograph.

Abstract: A high-efficiency, multiwavelength beam-expander optical system that employs dielectric-enhanced mirrors is disclosed. Each mirror includes a reflective multilayer coating formed from alternating layers of HfO2 and SiO2 that define, in order from the substrate surface, at least first and second sections, wherein the HfO2/SiO2 layer thicknesses are generally constant within a given section and get smaller section by section moving outward from the substrate surface. The first and second sections are respectively configured to optimally reflect different operating wavelengths so that the beam-expander optical system has an optical transmission of greater than 95% at the different operating wavelengths.

Abstract: A computer-implemented method and system are described for deblurring an image. The method may include accessing an image having a first blurred region and a second blurred region, and generating a first blur kernel for the first blurred region and a second blur kernel for the second blurred region. Thereafter, the first blur kernel is positioned with respect to the first blurred region, and the second blur kernel is positioned with respect to the second blurred region based on the position of the first blur kernel. The image is then deblurred by deconvolving the first blurred region with the first blur kernel, and deconvolving the second blurred region with the second blur kernel.

Abstract: A simulated tracking shot is generated from an image sequence in which a foreground feature moves relative to a background during capturing of the image sequence. The background is artificially blurred in the simulated tracking shot in a spatially-invariant manner corresponding to foreground motion relative to the background during a time span of the image sequence. The foreground feature can be substantially unblurred relative to a reference image selected from the image sequence. A system to generate the simulated tracking shot can be configured to derive spatially invariant blur kernels for a background portion by reconstructing or estimating a 3-D space of the captured scene, placing virtual cameras along a foreground trajectory in the 3-D space, and projecting 3-D background points on to the virtual cameras.

Abstract: Techniques for detecting and recognizing text may be provided. For example, an image may be analyzed to detect and recognize text therein. The analysis may involve detecting text components in the image. For example, multiple color spaces and multiple-stage filtering may be applied to detect the text components. Further, the analysis may involve extracting text lines based on the text components. For example, global information about the text components can be analyzed to generate best-fitting text lines. The analysis may also involve pruning and splitting the text lines to generate bounding boxes around groups of text components. Text recognition may be applied to the bounding boxes to recognize text therein.

Abstract: Aspects of the present disclosure relate to management of evaluated rule data sets. Specifically, a unreduced evaluated rule data set may contain a number of items to be compared or analyzed according to a number of rules, and may also contain the results of such analysis. An illustrative reduced evaluated data set can include the results of evaluated rules. When utilized in conjunction with an item data set and a rule data set, the information contained within the unreduced evaluated rule data set may be maintained. The reduce memory requirements of the reduced evaluated rule data set may facilitate storage of the reduced evaluated rule data set in faster to access memory, or may facilitate distributed computation of the reduced evaluated rule data set.

Abstract: Techniques are disclosed relating to lens modeling. In one embodiment, a lens model may be generated based on reference images of a pre-determined, known geometric pattern. The lens model may represent a spatially variant blur pattern across the image field of the lens used to capture the reference images. In one embodiment, the lens model may include Gaussian approximations of the blur that may minimize the difference between a location within a reference image and a corresponding location of a pre-determined, known geometric pattern. In one embodiment, the generated lens model may be applied to deblur a new image.