Patents by Inventor Felix Heide
Felix Heide has filed for patents to protect the following inventions. This listing includes patent applications that are pending as well as patents that have already been granted by the United States Patent and Trademark Office (USPTO).
-
Patent number: 10713537Abstract: A learning machine employs an image acquisition device for acquiring a set of training raw images. A processor determines a representation of a raw image, initializes a set of image representation parameters, defines a set of analysis parameters of an image analysis network configured to process the image's representation, and jointly trains the set of representation parameters and the set of analysis parameters to optimize a combined objective function. Processor executable instructions are organized into a module for transforming pixel-values of the raw image to produce a transformed image comprising pixels of variance-stabilized values, a module for successively performing processes of soft camera projection and image projection, and a module for inverse transforming the transformed pixels. The image projection process performs multi-level spatial convolution, pooling, subsampling, and interpolation.Type: GrantFiled: July 2, 2018Date of Patent: July 14, 2020Assignee: ALGOLUX INC.Inventor: Felix Heide
-
Publication number: 20200192287Abstract: A method for digital holography includes modeling a hologram using a forward propagation model that models propagation of a light field from a hologram plane to an image plane. The method further includes computing the hologram as a solution to an optimization problem that is based on the model. The method further includes configuring at least one spatial light modulator using the hologram. The method further includes illuminating the spatial light modulator using a light source to create a target image.Type: ApplicationFiled: December 11, 2019Publication date: June 18, 2020Inventors: Praneeth Kumar Chakravarthula, Felix Heide
-
Publication number: 20190154834Abstract: Systems and methods for imaging object velocity are provided. In an embodiment, at least one Time-of-Flight camera is used to capture a signal representative of an object in motion over an exposure time. Illumination and modulation frequency of the captured motion are coded within the exposure time. A change of illumination frequency is mapped to measured pixel intensities of the captured motion within the exposure time, and information about a Doppler shift in the illumination frequency is extracted to obtain a measurement of instantaneous per pixel velocity of the object in motion. The radial velocity information of the object in motion can be simultaneously captured for each pixel captured within the exposure time. In one or more aspects, the illumination frequency can be coded orthogonal to the modulation frequency of the captured motion. The change of illumination frequency can correspond to radial object velocity.Type: ApplicationFiled: August 5, 2016Publication date: May 23, 2019Applicants: KING ABDULLAH UNIVERSITY OF SCIENCE AND TECHNOLOGY, BONN UNIVERSITY, UNIVERSITY OF BRITISH COLUMBIA, STANFORD UNIVERSITY, BONN UNIVERSITYInventors: Wolfgang HEIDRICH, Felix HEIDE, Gordon WETZSTEIN, Matthias HULLIN
-
Publication number: 20190005360Abstract: A learning machine employs an image acquisition device for acquiring a set of training raw images. A processor determines a representation of a raw image, initializes a set of image representation parameters, defines a set of analysis parameters of an image analysis network configured to process the image's representation, and jointly trains the set of representation parameters and the set of analysis parameters to optimize a combined objective function. Processor executable instructions are organized into a module for transforming pixel-values of the raw image to produce a transformed image comprising pixels of variance-stabilized values, a module for successively performing processes of soft camera projection and image projection, and a module for inverse transforming the transformed pixels. The image projection process performs multi-level spatial convolution, pooling, subsampling, and interpolation.Type: ApplicationFiled: July 2, 2018Publication date: January 3, 2019Inventor: Felix HEIDE
-
Patent number: 9934714Abstract: System and method of displaying images in temporal superresolution by multiplicative superposition of cascaded display layers integrated in a display device. Using an original video with a target temporal resolution as a priori, a factorization process is performed to derive respective image data for presentation on each display layer. The multiple layers are refreshed in staggered intervals to synthesize a video with an effective refresh rate exceeding that of each individual display layer, e.g., by a factor equal to the number of layers. Further optically averaging neighboring pixels can minimize artifacts.Type: GrantFiled: March 17, 2015Date of Patent: April 3, 2018Assignee: NVIDIA CORPORATIONInventors: Felix Heide, Douglas Lanman, Dikpal Reddy, Jan Kautz, Kari Pulli, David Luebke
-
Patent number: 9905196Abstract: A computer implemented method of determining a latent image from an observed image is disclosed. The method comprises implementing a plurality of image processing operations within a single optimization framework, wherein the single optimization framework comprises solving a linear minimization expression. The method further comprises mapping the linear minimization expression onto at least one non-linear solver. Further, the method comprises using the non-linear solver, iteratively solving the linear minimization expression in order to extract the latent image from the observed image, wherein the linear minimization expression comprises: a data term, and a regularization term, and wherein the regularization term comprises a plurality of non-linear image priors.Type: GrantFiled: September 26, 2016Date of Patent: February 27, 2018Assignee: NVIDIA CORPORATIONInventors: Dawid Stanislaw Pajak, Felix Heide, Nagilla Dikpal Reddy, Mushfiqur Rouf, Jan Kautz, Kari Pulli, Orazio Gallo
-
Patent number: 9892669Abstract: System and method of displaying images in spatial/temporal superresolution by multiplicative superposition of cascaded display layers integrated in a display device. Using an original image with a target spatial/temporal resolution as a priori, a factorization process is performed to derive respective image data for presentation on each display layer. The cascaded display layers may be progressive and laterally shifted with each other, resulting in an effective spatial resolution exceeding the native display resolutions of the display layers. Factorized images may be refreshed on respective display layers in synchronization or out of synchronization.Type: GrantFiled: March 17, 2015Date of Patent: February 13, 2018Assignee: Nvidia CorporationInventors: Felix Heide, Douglas Lanman, Dikpal Reddy, Jan Kautz, Kari Pulli, David Luebke
-
Patent number: 9558712Abstract: A computer implemented method of determining a latent image from an observed image is disclosed. The method comprises implementing a plurality of image processing operations within a single optimization framework, wherein the single optimization framework comprises solving a linear minimization expression. The method further comprises mapping the linear minimization expression onto at least one non-linear solver. Further, the method comprises using the non-linear solver, iteratively solving the linear minimization expression in order to extract the latent image from the observed image, wherein the linear minimization expression comprises: a data term, and a regularization term, and wherein the regularization term comprises a plurality of non-linear image priors.Type: GrantFiled: January 20, 2015Date of Patent: January 31, 2017Assignee: NVIDIA CORPORATIONInventors: Dawid Stanislaw Pajak, Felix Heide, Nagilla Dikpal Reddy, Mushfiqur Rouf, Jan Kautz, Kari Pulli, Orazio Gallo
-
Publication number: 20170011710Abstract: A computer implemented method of determining a latent image from an observed image is disclosed. The method comprises implementing a plurality of image processing operations within a single optimization framework, wherein the single optimization framework comprises solving a linear minimization expression. The method further comprises mapping the linear minimization expression onto at least one non-linear solver. Further, the method comprises using the non-linear solver, iteratively solving the linear minimization expression in order to extract the latent image from the observed image, wherein the linear minimization expression comprises: a data term, and a regularization term, and wherein the regularization term comprises a plurality of non-linear image priors.Type: ApplicationFiled: September 26, 2016Publication date: January 12, 2017Inventors: Dawid Stanislaw Pajak, Felix Heide, Nagilla Dikpal Reddy, Mushfiqur Rouf, Jan Kautz, Kari Pulli, Orazio Gallo
-
Patent number: 9343020Abstract: In exemplary implementations of this invention, light from a backlight is transmitted through two stacked LCDs and then through a diffuser. The front side of the diffuser displays a time-varying sequence of 2D images. Processors execute an optimization algorithm to compute optimal pixel states in the first and second LCDs, respectively, such that for each respective image in the sequence, the optimal pixel states minimize, subject to one or more constraints, a difference between a target image and the respective image. The processors output signals to control actual pixel states in the LCDs, based on the computed optimal pixel states. The 2D images displayed by the diffuser have a higher spatial resolution than the native spatial resolution of the LCDs. Alternatively, the diffuser may be switched off, and the device may display either (a) 2D images with a higher dynamic range than the LCDs, or (b) an automultiscopic display.Type: GrantFiled: August 5, 2014Date of Patent: May 17, 2016Assignee: Massachusetts Institute of TechnologyInventors: Felix Heide, Gordon Wetzstein, James Gregson, Ramesh Raskar, Wolfgang Heidrich
-
Patent number: 9307208Abstract: There is provided a computer-implemented method for solving inverse imaging problems to compensate for distortions in an image. The method comprises: minimizing a cost objective function containing a data fitting term and one or more image prior terms to each of the plurality of channels, the one or more image prior terms comprising cross-channel information for a plurality of channels derived from the image.Type: GrantFiled: March 7, 2014Date of Patent: April 5, 2016Assignee: THE UNIVERSITY OF BRITISH COLUMBIAInventors: Wolfgang Heidrich, Felix Heide, Mushfiqur Rouf, Matthias Hullin
-
Publication number: 20150310789Abstract: System and method of displaying images in spatial/temporal superresolution by multiplicative superposition of cascaded display layers integrated in a display device. Using an original image with a target spatial/temporal resolution as a priori, a factorization process is performed to derive respective image data for presentation on each display layer. The cascaded display layers may be progressive and laterally shifted with each other, resulting in an effective spatial resolution exceeding the native display resolutions of the display layers. Factorized images may be refreshed on respective display layers in synchronization or out of synchronization.Type: ApplicationFiled: March 17, 2015Publication date: October 29, 2015Inventors: Felix Heide, Douglas Lanman, Dikpal Reddy, Jan Kautz, Kari Pulli, David Luebke
-
Publication number: 20150310798Abstract: System and method of displaying images in temporal superresolution by multiplicative superposition of cascaded display layers integrated in a display device. Using an original video with a target temporal resolution as a priori, a factorization process is performed to derive respective image data for presentation on each display layer. The multiple layers are refreshed in staggered intervals to synthesize a video with an effective refresh rate exceeding that of each individual display layer, e.g., by a factor equal to the number of layers. Further optically averaging neighboring pixels can minimize artifacts.Type: ApplicationFiled: March 17, 2015Publication date: October 29, 2015Inventors: Felix Heide, Douglas Lanman, Dikpal Reddy, Jan Kautz, Kari Pulli, David Luebke
-
Publication number: 20150254810Abstract: There is provided a computer-implemented method for solving inverse imaging problems to compensate for distortions in an image. The method comprises: minimizing a cost objective function containing a data fitting term and one or more image prior terms to each of the plurality of channels, the one or more image prior terms comprising cross-channel information for a plurality of channels derived from the image.Type: ApplicationFiled: March 7, 2014Publication date: September 10, 2015Applicant: The University of British ColumbiaInventors: Wolfgang Heidrich, Felix Heide, Mushfiqur Rouf, Matthias Hullin
-
Publication number: 20150206504Abstract: A computer implemented method of determining a latent image from an observed image is disclosed. The method comprises implementing a plurality of image processing operations within a single optimization framework, wherein the single optimization framework comprises solving a linear minimization expression. The method further comprises mapping the linear minimization expression onto at least one non-linear solver. Further, the method comprises using the non-linear solver, iteratively solving the linear minimization expression in order to extract the latent image from the observed image, wherein the linear minimization expression comprises: a data term, and a regularization term, and wherein the regularization term comprises a plurality of non-linear image priors.Type: ApplicationFiled: January 20, 2015Publication date: July 23, 2015Inventors: Dawid Stanislaw Pajak, Felix Heide, Nagilla Dikpal Reddy, Mushfiqur Rouf, Jan Kautz, Kari Pulli, Orazio Gallo
-
Publication number: 20150035880Abstract: In exemplary implementations of this invention, light from a backlight is transmitted through two stacked LCDs and then through a diffuser. The front side of the diffuser displays a time-varying sequence of 2D images. Processors execute an optimization algorithm to compute optimal pixel states in the first and second LCDs, respectively, such that for each respective image in the sequence, the optimal pixel states minimize, subject to one or more constraints, a difference between a target image and the respective image. The processors output signals to control actual pixel states in the LCDs, based on the computed optimal pixel states. The 2D images displayed by the diffuser have a higher spatial resolution than the native spatial resolution of the LCDs. Alternatively, the diffuser may be switched off, and the device may display either (a) 2D images with a higher dynamic range than the LCDs, or (b) an automultiscopic display.Type: ApplicationFiled: August 5, 2014Publication date: February 5, 2015Applicant: MASSACHUSETTS INSTITUTE OF TECHNOLOGYInventors: Felix Heide, Gordon Wetzstein, James Gregson, Ramesh Raskar, Wolfgang Heidrich