Abstract: One exemplary embodiment involves receiving an image with a set of undefined pixels and a set of defined pixels and recursively modifying the image to generate a seamless composition comprising only defined pixels. Disclosed are embodiments for recursively modifying the image by recursively down sampling the image by a factor to generate a plurality of down sampled images until the down sampled image generated at each recursive down sampling lacks undefined pixels and then recursively up sampling each one of the down sampled images by the factor to generate an up sampled image from the respective down sampled image. Additionally, at each recursive up sampling instance, pasting the next recursively occurring down sampled image on the up sampled image to generate the next recursively occurring image for up sampling.

Abstract: A system and method are proposed for performing gradient-based compositing in a manner that reduces the appearance of residual errors as compared to traditional techniques. In embodiments, a target gradient field may be determined for a composite image and a solution composite image may be sought which minimizes a weighted aggregate error between the gradient field of the composite image and the target gradient field. The weighted aggregate error may be calculated such that a greater weight is placed on residuals occurring in smooth regions than on those occurring in textured regions, since residuals in textured regions may be less noticeable. In some embodiments, a weighted aggregate error may be defined as a weighted least-squares sum and/or minimized by solving a weighted Poisson equation. The computational cost of solving the weighted Poisson equation may be mitigated by ensuring that the equation is linear.

Abstract: A computer-implemented method includes segmenting a plurality of video frames of a sequence of video frames into a first portion that includes a selected visual object represented in the video frame and a second portion that includes a background represented in the video frame. The selected visual object is selected by using a selection envelope.

Abstract: Techniques are disclosed relating to automatically adjusting images. In one embodiment, an image may be automatically adjusted based on a regression model trained with a database of raw and adjusted images. In one embodiment, an image may be automatically adjusted based on a model trained by both a database of raw and adjusted images and a small set of images adjusted by a different user. In one embodiment, an image may be automatically adjusted based on a model trained by a database of raw and adjusted images and predicted differences between a user's adjustment to a small set of images and a predicted adjustment based on the database of raw and adjusted images.

Abstract: A system and method for a blur brush performing adaptive bilateral filtering is disclosed. The method may include receiving user input selecting an area of an image to be filtered, such as by pointing to the image area using the blur brush. The selected image may comprise an edge and a plurality of pixels. The method may operate to the blur brush identifying the edge in the selected image area. The method may operate to apply a filter tool (e.g., a bilateral filter) to the selected image area, while preserving the edge. The methods may be implemented by program instructions executing in parallel on CPU(s) or GPUs.

Abstract: A computer-implemented method includes segmenting an input image into a plurality of image cues, each image cue representing a unique set of pixels of the input image. For each image cue, the method includes determining a set of image components, wherein each image component is associated with at least one adjustable factor to represent at least one characteristic of the image cue.

Abstract: A system and method are proposed for performing gradient-based compositing in a manner that reduces the appearance of residual errors as compared to traditional techniques. In embodiments, a target gradient field may be determined for a composite image and a solution composite image may be sought which minimizes a weighted aggregate error between the gradient field of the composite image and the target gradient field. The weighted aggregate error may be calculated such that a greater weight is placed on residuals occurring in smooth regions than on those occurring in textured regions, since residuals in textured regions may be less noticeable. In some embodiments, a weighted aggregate error may be defined as a weighted least-squares sum and/or minimized by solving a weighted Poisson equation. The computational cost of solving the weighted Poisson equation may be mitigated by ensuring that the equation is linear.

Abstract: Techniques are disclosed relating to automatically adjusting images. In one embodiment, an image may be automatically adjusted based on a regression model trained with a database of raw and adjusted images. In one embodiment, an image may be automatically adjusted based on a model trained by both a database of raw and adjusted images and a small set of images adjusted by a different user. In one embodiment, an image may be automatically adjusted based on a model trained by a database of raw and adjusted images and predicted differences between a user's adjustment to a small set of images and a predicted adjustment based on the database of raw and adjusted images.

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.

Abstract: Various embodiments of a system and methods for real-time image collection and sharing are described. A group of geographically co-located mobile device users may capture images on the mobile devices during a session. The devices may send the images, during the same session, to a database where the images may be stored as a real-time image collection. The image collection may be shared, during the same session, among the mobile device users. The users may receive notification of new images in the image collection based on user preferences. The users may request, from the image collection, images containing particular content by identifying the content on an image and sending a search request which includes the identified content. A computation server may maintain a data set which associates similar images in the image collection. The data set may be updated for each new image that is added to the image collection.

Abstract: Various embodiments of a system and methods for real-time image collection and sharing are described. A group of geographically co-located mobile device users may capture images on the mobile devices during a session. The devices may send the images, during the same session, to a database where the images may be stored as a real-time image collection. The image collection may be shared, during the same session, among the mobile device users. The users may receive notification of new images in the image collection based on user preferences. The users may request, from the image collection, images containing particular content by identifying the content on an image and sending a search request which includes the identified content. A computation server may maintain a data set which associates similar images in the image collection. The data set may be updated for each new image that is added to the image collection.

Abstract: Techniques are disclosed relating to modifying an automatically predicted adjustment. In one embodiment, the automatically predicted adjustment may be adjusted, for example, based on a rule. The automatically predicted adjustment may be based on a machine learning prediction. A new image may be globally adjusted based on the modified automatically predicted adjustment.

Abstract: A system and method for a blur brush performing adaptive bilateral filtering is disclosed. The method may include receiving user input selecting an area of an image to be filtered, such as by pointing to the image area using the blur brush. The selected image may comprise an edge and a plurality of pixels. The method may operate to the blur brush identifying the edge in the selected image area. The method may operate to apply a filter tool (e.g., a bilateral filter) to the selected image area, while preserving the edge. The methods may be implemented by program instructions executing in parallel on CPU(s) or GPUs.

Abstract: Methods, apparatus, and computer-readable storage media for tone mapping High Dynamic Range (HDR) images. An input HDR image is separated into luminance and color. Luminance is processed to obtain a base layer and a detail layer. The base layer is compressed according to a non-linear remapping function to reduce the dynamic range, and the detail layer is adjusted. The layers are combined to generate output luminance, and the output luminance and color are combined to generate an output image. A base layer compression technique may be used that analyzes the details and compresses the base layer accordingly to provide space at the top of the intensity scale where the details are displayed to thus generate output images that are visually better than images generated using conventional techniques. User interface elements may be provided via which a user may control one or more parameters of the tone mapping method.

Abstract: A computer-implemented method that includes segmenting a training image into training image patches, where each training image patch is represented by a linear combination of dictionary image patches from an image dictionary, and each dictionary image patch has a sparse representation coefficient. The method includes segmenting a stylized training image into stylized training image patches, where each stylized training image patch is represented by a linear combination of stylized dictionary image patches from a stylized image dictionary, and each stylized dictionary image patch has a sparse representation coefficient. The method also includes training the image dictionary with the training image patches and the stylized image dictionary with the stylized training image patches in a substantially simultaneous manner. The sparse representation coefficient for each training image patch is substantially similar to the sparse representation coefficient for the corresponding stylized training image patch.

Abstract: Methods, apparatus, and computer-readable storage media for tone mapping High Dynamic Range (HDR) images. The HDR image is separated into luminance and color. Luminance is processed according to the parameters to obtain a base layer and a detail layer. The base layer is compressed into a lower dynamic range and the detail layer is adjusted according to the parameters. The compressed base layer, the detail layer, and the color component may be output as separate layers, and various image processing tools and techniques may be applied to the component layers separately to modify the layer(s). One or more tone-mapped images may be generated by merging the modified layers. Thus, each layer of the tone-mapped image may be processed separately using various image processing tools or techniques to modify the output of the tone mapping technique in a wide variety of ways.

Abstract: The present invention provides two-scale tone management of an input image. A method of providing two-scale tone management contains the steps of: separating the input image into a base layer and a detail layer; separating a model image into a base layer and a detail layer; analyzing the input image globally for global contrast; analyzing the input image locally for local contrast; and performing detail preservation of the input image. A system contains a memory and a processor, where the processor is configured by the memory to perform the steps of: separating the input image into a base layer and a detail layer; separating a model image into a base layer and a detail layer; analyzing the input image globally for global contrast; analyzing the input image locally for local contrast; and performing detail preservation of the input image.

Abstract: The present invention provides gradient preservation of an input image. A method of providing gradient preservation of an input image, comprises the steps of: computing gradient fields of the input image; computing gradient fields of a remapped image, wherein the remapped image is the input image remapped; comparing a gradient of the remapped image to a predefined range of acceptable values, wherein the acceptable values depend on the input gradient; changing values of the gradient of the remapped image if the values of the gradient of the remapped image are not within the predefined range of acceptable values depending on the input gradient, resulting in x and y components of a modified gradient; and reconstructing the input image using the x and y components of the modified gradient.

Abstract: A computer-implemented method includes segmenting an input image into a plurality of image cues, each image cue representing a unique set of pixels of the input image. For each image cue, the method includes determining a set of image components, wherein each image component is associated with at least one adjustable factor to represent at least one characteristic of the image cue.

Abstract: A computer-implemented method includes segmenting a plurality of video frames of a sequence of video frames into a first portion that includes a selected visual object represented in the video frame and a second portion that includes a background represented in the video frame. The selected visual object is selected by using a selection envelope.