Abstract: Systems, methods and computer-readable storage media are disclosed for accelerating bitmap remoting by extracting patterns from source bitmaps. A server takes a source image, and performs an edge-detection operation on it. From this edge-detected image, connected segments of the image are determined by executing multiple iterations of a small operation upon the image in parallel—for instance, by assigning each non-white pixel a unique value, then assigning each pixel the minimum value among itself and its neighbors until no pixel is assigned a new value in an iteration. Executing these operations in parallel greatly reduces the time required to identify the connected segments. When the segments are identified, they may be cached by the client so that they do not need to be re-sent to the client when re-encountered by the server.

Abstract: Systems, methods and computer-readable storage media are disclosed for accelerating bitmap remoting by extracting non-grid tiles from source bitmaps. A server takes a source image, identifies possibly repetitive features, and tiles the image. For each tile that contains part of a possibly repetitive feature, the server replaces that part with the dominant color of the tile. The system then sends to a client a combination of new tiles and features, and indications to tiles and features that the client has previously received and stored, along with an indication of how to recreate the image based on the tiles and features.

Abstract: Described herein are various technologies for generating descriptors for image patches. An image patch can be received, and gradients of pixels in the image patch can be determined. The gradients are normalized based upon an average magnitude of the gradients in a local spatial region with respect to a given pixel under consideration. A four-dimensional histogram is defined that takes into consideration pixel orientation, and normalized gradients are selectively assigned to bins of the histogram. The bins are binarized as a function of a number of gradients assigned thereto, and the binarized bins can be utilized as a descriptor for the image patch.

Abstract: Technology is described for detecting an interest point in an image using edges. An example method can include the operation of computing locally normalized edge magnitudes and edge orientations for the image using a processor to form a normalized gradient image. The normalized gradient image can be divided into a plurality of image orientation maps having edge orientations. Orientation dependent filtering can be applied to the image orientation maps to form response images. A further operation can be summing the response images to obtain an aggregated filter response image. Maxima can be identified in spatial position and scale in the aggregated filter response image. Maxima in the aggregated filter response image can be defined as interest points.

Abstract: The claimed subject matter provides for systems and/or methods for identification of instances of an object of interest in 2D images by creating a database of 3D curve models of each desired instance and comparing an image of an object of interest against such 3D curve models of instances. The present application describes identifying and verifying the make and model of a car from a possibly single image—after the models have been populated with training data of test images of many makes and models of cars. In one embodiment, an identification system may be constructed by generating a 3D curve model by back-projecting edge points onto a visual hull reconstruction from silhouettes of an instance. The system and methods employ chamfer distance and orientation distance provides reasonable verification performance, as well as an appearance model for the taillights of the car to increase the robustness of the system.

Abstract: Methods and systems for generating a depth map are provided. The method includes projecting an infrared (IR) dot pattern onto a scene. The method also includes capturing stereo images from each of two or more synchronized IR cameras, detecting a number of dots within the stereo images, computing a number of feature descriptors for the dots in the stereo images, and computing a disparity map between the stereo images. The method further includes generating a depth map for the scene using the disparity map.

Abstract: Methods and systems for generating free viewpoint video using an active infrared (IR) stereo module are provided. The method includes computing a depth map for a scene using an active IR stereo module. The depth map may be computed by projecting an IR dot pattern onto the scene, capturing stereo images from each of two or more synchronized IR cameras, detecting dots within the stereo images, computing feature descriptors corresponding to the dots in the stereo images, computing a disparity map between the stereo images, and generating the depth map using the disparity map. The method also includes generating a point cloud for the scene using the depth map, generating a mesh of the point cloud, and generating a projective texture map for the scene from the mesh of the point cloud. The method further includes generating the video for the scene using the projective texture map.

Abstract: Techniques are disclosed for acceleration techniques for improved image remoting. A rolling 2D hash of a first image sent to a client is computed. When the server has a second image to send to the client, it calculates a rolling 2D hash of the new image. It also calculates “pivot points” for the images based on the rolling 2D hashes. Based on the pivot points, it determines possible matching hash windows between the two images that correspond to window moves or scrolls. Where a match is confirmed, it determines whether a “larger” a larger matching rectangle exists between the two images. It then instructs the client to display the matching rectangle that exists in the first image that the client has in the appropriate location in the second image, thereby saving the bandwidth requirements to re-transmit it to the client.

Abstract: A mobile device having the capability of performing real-time location recognition with assistance from a server is provided. The approximate geophysical location of the mobile device is uploaded to the server. Based on the mobile device's approximate geophysical location, the server responds by sending the mobile device a message comprising a classifier and a set of feature descriptors. This can occur before an image is captured for visual querying. The classifier and feature descriptors are computed during an offline training stage using techniques to minimize computation at query time. The classifier and feature descriptors are used to perform visual recognition in real-time by performing the classification on the mobile device itself.

Abstract: The stereo movie editing technique described herein combines knowledge of both multi-view stereo algorithms and human depth perception. The technique creates a digital editor, specifically for stereographic cinema. The technique employs an interface that allows intuitive manipulation of the different parameters in a stereo movie setup, such as camera locations and screen position. Using the technique it is possible to reduce or enhance well-known stereo movie effects such as cardboarding and miniaturization. The technique also provides new editing techniques such as directing the user's attention and easier transitions between scenes.

Abstract: Architecture that guides the freeform drawing of objects by a user to enable the user to produce improved drawings without significant training. As the user draws, the architecture dynamically updates a relevant shadow image proximate (e.g., underlying) the user's strokes. The strokes overlay an evolving shadow image, which shadow image is suggestive of object contours that guide the user during the drawing process. Relevant edge images selected from a large database are automatically blended to construct the shadow image. As the user draws, the strokes are dynamically analyzed using an encoding of overlapping windows for fast matching with the database of images. A top ranked set of matching database edge images are aligned to the drawing, a set of spatially varying weights blend the edge images into the shadow image, and a scoring technique is employed to select the optimum shadow image for display.

Abstract: Technology is described for detecting an interest point in an image using edges. An example method can include the operation of computing locally normalized edge magnitudes and edge orientations for the image using a processor to form a normalized gradient image. The normalized gradient image can be divided into a plurality of image orientation maps having edge orientations. Orientation dependent filtering can be applied to the image orientation maps to form response images. A further operation can be summing the response images to obtain an aggregated filter response image. Maxima can be identified in spatial position and scale in the aggregated filter response image. Maxima in the aggregated filter response image can be defined as interest points.

Abstract: Video signatures are generated when a small change in video information between consecutive frames N?1 and N is followed by a large change in video information between consecutive frames N and N+1. Information from frames N and/or N+1 is used to form a video signature.

Abstract: Described is a technology in which video shots are clustered based upon the location at which the shots were captured. A global energy function is optimized, including a first term that computes clusters so as to be reasonably dense and well connected, to match the possible shots that are captured at a location, e.g., based on similarity scores between pairs of shots. A second term is a temporal prior that encourages subsequent shots to be placed in the same cluster. The shots may be represented as nodes of a minimum spanning tree having edges with weights that are based on the similarity score between the shots represented by their respective nodes. Agglomerative clustering is performed by selecting pairs of available clusters, merging the pairs and keeping the pair with the lowest cost. Clusters are iteratively merged until a stopping criterion or criteria is met (e.g., only a single cluster remains).

Abstract: A process for compressing and decompressing non-keyframes in sequential sets of contemporaneous video frames making up multiple video streams where the video frames in a set depict substantially the same scene from different viewpoints. Each set of contemporaneous video frames has a plurality frames designated as keyframes with the remaining being non-keyframes. In one embodiment, the non-keyframes are compressed using a multi-directional spatial prediction technique. In another embodiment, the non-keyframes of each set of contemporaneous video frames are compressed using a combined chaining and spatial prediction compression technique. The spatial prediction compression technique employed can be a single direction technique where just one reference frame, and so one chain, is used to predict each non-keyframe, or it can be a multi-directional technique where two or more reference frames, and so chains, are used to predict each non-keyframe.

Abstract: Described herein are various technologies for generating descriptors for image patches. An image patch can be received, and gradients of pixels in the image patch can be determined. The gradients are normalized based upon an average magnitude of the gradients in a local spatial region with respect to a given pixel under consideration. A four-dimensional histogram is defined that takes into consideration pixel orientation, and normalized gradients are selectively assigned to bins of the histogram. The bins are binarized as a function of a number of gradients assigned thereto, and the binarized bins can be utilized as a descriptor for the image patch.

Abstract: A two-dimensional blur kernel is computed for a digital image by first estimating a sharp image from the digital image. The sharp image is derived from the digital image by sharpening at least portions of the digital image. The two-dimensional blur function is computed by minimizing an optimization algorithm that estimates the blur function.

Abstract: Techniques are disclosed for acceleration techniques for improved image remoting. A rolling 2D hash of a first image sent to a client is computed. When the server has a second image to send to the client, it calculates a rolling 2D hash of the new image. It also calculates “pivot points” for the images based on the rolling 2D hashes. Based on the pivot points, it determines possible matching hash windows between the two images that correspond to window moves or scrolls. Where a match is confirmed, it determines whether a “larger” a larger matching rectangle exists between the two images. It then instructs the client to display the matching rectangle that exists in the first image that the client has in the appropriate location in the second image, thereby saving the bandwidth requirements to re-transmit it to the client.

Abstract: Systems, methods and computer-readable storage media are disclosed for accelerating bitmap remoting by extracting patterns from source bitmaps. A server takes a source image, and performs an edge-detection operation on it. From this edge-detected image, connected segments of the image are determined by executing multiple iterations of a small operation upon the image in parallel—for instance, by assigning each non-white pixel a unique value, then assigning each pixel the minimum value among itself and its neighbors until no pixel is assigned a new value in an iteration. Executing these operations in parallel greatly reduces the time required to identify the connected segments. When the segments are identified, they may be cached by the client so that they do not need to be re-sent to the client when re-encountered by the server.

Abstract: Systems, methods and computer-readable storage media are disclosed for accelerating bitmap remoting by extracting non-grid tiles from source bitmaps. A server takes a source image, identifies possibly repetitive features, and tiles the image. For each tile that contains part of a possibly repetitive feature, the server replaces that part with the dominant color of the tile. The system then sends to a client a combination of new tiles and features, and indications to tiles and features that the client has previously received and stored, along with an indication of how to recreate the image based on the tiles and features.