Abstract: Techniques related to generating a virtual view from multi-view images for presentation to a viewer are discussed. Such techniques include determining, based on a viewer position relative to a display region, first and second crop positions of planar image and cropping the planar image to a cropped planar image to fill the display region using the first and second crop positions such that the first and second crop positions define an asymmetric frustum between the cropped planar image and a virtual window corresponding to the display region.

Abstract: A mechanism is described for facilitating cinematic space-time view synthesis in computing environments according to one embodiment. A method of embodiments, as described herein, includes capturing, by one or more cameras, multiple images at multiple positions or multiple points in times, where the multiple images represent multiple views of an object or a scene, where the one or more cameras are coupled to one or more processors of a computing device. The method further includes synthesizing, by a neural network, the multiple images into a single image including a middle image of the multiple images and representing an intermediary view of the multiple views.

Abstract: A plurality of frames of a video recorded by a video camera and depth maps of the plurality of frames are stored in a data storage. One or more target video camera positions are determined. Each frame of the plurality of frames is associated with one or more of the target video camera positions. For each frame, one or more synthesized frames from the viewpoint of the one or more target camera positions associated with that frame are generated by applying a view interpolation algorithm to that frame using the color pixels of that frame and the depth map of that frame. Users can provide their input about the new camera positions and other camera parameters through multiple input modalities. The synthesized frames are concatenated to create a modified video. Other embodiments are also described and claimed.

Abstract: Techniques related to interpolating an intermediate view image from multi-view images are discussed. Such techniques include downsampling first and second images that represent a view of a scene, generating a disparity map based on applying a first CNN to the downscaled first and second images, translating the downscaled first and second images using the disparity map, applying a second CNN to the translated downscaled first and second images and the disparity map to generate a downscaled intermediate image, and upscaling the downscaled intermediate image to an intermediate image at the resolution of the first and second images using an image super-resolution convolutional neural network.

Abstract: Techniques related to interpolating an intermediate view image from multi-view images are discussed. Such techniques include downsampling first and second images that represent a view of a scene, generating a disparity map based on applying a first CNN to the downscaled first and second images, translating the downscaled first and second images using the disparity map, applying a second CNN to the translated downscaled first and second images and the disparity map to generate a downscaled intermediate image, and upscaling the downscaled intermediate image to an intermediate image at the resolution of the first and second images using an image super-resolution convolutional neural network.

Abstract: Disclosed herein is display stacks for virtual reality (VR) displays arranged to present an image where a focal point of the image is presented in higher resolution than a periphery around the focal point. The disclosure provides systems and methods for presenting an image comprising a high resolution having a smaller field of view than that with which a lower resolution image is presented around a periphery of the high resolution image.

Abstract: Techniques related to multiview video stabilization are discussed. Such techniques may include performing single view video stabilization on a reference video stream and propagating the single view video stabilization to another video stream based on a target disparity map.

Abstract: An example method is described herein. The method includes executing dense feature matching on an image pair that is down sampled to obtain a first set of feature correspondences for each pixel of the down sampled image pair. The method also includes calculating a neighborhood correspondence based on the first set of feature correspondences for each pixel in a first image of the image pair. Further, the method includes executing sparse feature matching on stereoscopic patch pairs from the image pair based on the neighborhood correspondence for each pixel to obtain correspondence estimates for each stereoscopic patch pair. Finally, the method includes refining the correspondence estimates for each stereoscopic patch pair to obtain a semi-dense set of feature correspondences by applying a geometric constraint to the correspondence estimates and retaining correspondences that satisfy the geometric constraint.

Abstract: Techniques related to generating a virtual view of a scene from a position between positions of known input images for presentation to a viewer are discussed. Such techniques include applying a convolution mask that approximates an inverse of a linear combination of at least the horizontal and vertical convolution matrices representative of gradient detection in the input images to a virtual intermediate gradient image that is a combination of at least an interpolated virtual image at the position, an interpolated horizontal gradient map at the position, and an interpolated vertical gradient map at the position to generate a final virtual image for the position.

Abstract: Techniques related to feature detection and matching fisheye images are discussed. Such techniques include determining a geometric constraint for the feature matching using match results from a first image based feature matching and generating resultant matches based on applying the geometric constraint and a second image based feature matching that applies a looser image based feature matching requirement than the image based feature matching.

Abstract: Techniques are provided for perception enhancement of light fields (LFs) for use in integral display applications. A methodology implementing the techniques according to an embodiment includes receiving one or more LF views and a disparity map associated with each LF view. The method also includes quantizing the disparity map into planes, where each plane is associated with a selected range of depth values. The method further includes slicing the LF view into layers, where each layer comprises pixels of the LF view associated with one of the planes. The method further includes shifting each of the layers in a lateral direction by an offset distance. The offset distance is based on a viewing angle associated with the LF view and further based on the depth values of the associated plane. The method also includes merging the shifted layers to generate a synthesized LF view with increased parallax.

Abstract: An example system for generating light field content is described herein. The system includes a receiver to receive a plurality of images and a calibrator to intrinsically calibrate a camera. The system also includes a corrector and projector undistort the images and project the undistorted images to generate undistorted rectilinear images. An extrinsic calibrator may rectify and align the undistorted rectilinear images. Finally, the system includes a view interpolator to perform intermediate view interpolation on the rectified and aligned undistorted rectilinear images.

Abstract: An example system for generating stereoscopic light field panoramas includes a receiver to receive a plurality of synchronized images. The system also includes a calibrator and projector to calibrate the synchronized images, undistort the synchronized images, and project the undistorted images to a sphere to generate undistorted rectilinear images. The system further includes a disparity estimator to estimate a disparity between neighboring views of the undistorted rectilinear images to determine an optical flow between the undistorted rectilinear images. The system includes a view interpolator to perform in-between view interpolation on the undistorted rectilinear images based on the optical flow. The system also further includes a light field panorama generator to generate a stereoscopic light field panorama for a plurality of perspectives using concentric viewing circles.

Abstract: A mechanism is described for facilitating cinematic space-time view synthesis in computing environments according to one embodiment. A method of embodiments, as described herein, includes capturing, by one or more cameras, multiple images at multiple positions or multiple points in times, where the multiple images represent multiple views of an object or a scene, where the one or more cameras are coupled to one or more processors of a computing device. The method further includes synthesizing, by a neural network, the multiple images into a single image including a middle image of the multiple images and representing an intermediary view of the multiple views.

Abstract: Technology for improving performance of a personal display device by variably controlling the emission divergence and/or the emission direction of light from the pixels of the display as a function of the location of the pixel within a display.

Abstract: Extended depth of focus integral displays are disclosed. An example integral display includes a display screen to display an image including a plurality of interlaced elemental images that represent different views of a three-dimensional (3D) image, and an array of lenses proximate the display to integrate the elemental images to form the 3D image, the lenses selectively switchable between a first focal length and a second focal length to increase a depth of focus of the 3D image. Another example integral display includes a display screen to display an image including a plurality of interlaced elemental images that represent different views of a 3D image, and an array of lenses proximate the display to integrate the elemental images to form the 3D image, the array of lenses including first lenses having a first focal length interlaced with second lenses having a second focal length.

Abstract: Techniques related to interpolating an intermediate view image from multi-view images are discussed. Such techniques include downsampling first and second images that represent a view of a scene, generating a disparity map based on applying a first CNN to the downscaled first and second images, translating the downscaled first and second images using the disparity map, applying a second CNN to the translated downscaled first and second images and the disparity map to generate a downscaled intermediate image, and upscaling the downscaled intermediate image to an intermediate image at the resolution of the first and second images using an image super-resolution convolutional neural network.

Abstract: Techniques related to generating a virtual view from multi-view images for presentation to a viewer are discussed. Such techniques include determining, based on a viewer position relative to a display region, first and second crop positions of planar image and cropping the planar image to a cropped planar image to fill the display region using the first and second crop positions such that the first and second crop positions define an asymmetric frustum between the cropped planar image and a virtual window corresponding to the display region.

Abstract: A system for robust disparity estimation in the presence of significant intensity variations for camera arrays is described herein. The system comprises a camera array, a memory and a processor. The memory is configured to store imaging data. The processor is coupled to the memory and the camera array. When executing instructions, the processor is to obtain a plurality of images and generate a sequence of color matched images, wherein the sequence includes each pair of images in the plurality of images. The processor is also to calculate a plurality of disparity points based on the sequence of color matched images.

Abstract: A mechanism is described for facilitating real-time capturing, processing, and rendering of data according to one embodiment. A method of embodiments, as described herein, includes facilitating a capturing device to capture data of a scene, where the data includes a video having at least one of a two-and-a-half-dimensional video (2.5D) or a three-dimensional (3D) video. The method may further include processing, in real-time, the data to generate contents representing a 3D rendering of the data, and facilitating a display device to render, in real-time, the contents.