Patents by Inventor Peter N. Belhumeur
Peter N. Belhumeur 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).
-
Publication number: 20190286288Abstract: One or more embodiments of a digital content system allow a user to conveniently search and/or navigate through a collection of digital content items. In particular, a user can interact with a client device to search for and identify one or more digital content items within a collection of digital content items. For example, the digital content system may provide a photo from a collection of photos via a graphical user interface. The digital content system can receive a user input identifying a selection of one or more visual features within the photo. Based on the selected visual feature(s), the digital content system may identify photos from the collection of photos that include the identified visual feature(s) and provide access to the identified photos via the graphical user interface.Type: ApplicationFiled: June 6, 2019Publication date: September 19, 2019Inventors: Peter N. Belhumeur, David J. Kriegman, Thomas Berg
-
Patent number: 10417321Abstract: The present disclosure is directed toward systems and methods to quickly and accurately identify boundaries of a displayed document in a live camera image feed, and provide a document boundary indicator within the live camera image feed. For example, systems and methods described herein utilize different display document detection processes in parallel to generate and provide a document boundary indicator that accurately corresponds with a displayed document within a live camera image feed. Thus, a user of the mobile computing device can easily see whether the document identification system has correctly identified the displayed document within the camera viewfinder feed.Type: GrantFiled: July 24, 2017Date of Patent: September 17, 2019Assignee: DROPBOX, INC.Inventors: Nils Peter Welinder, Peter N. Belhumeur, Ying Xiong, Jongmin Baek, Simon Kozlov, Thomas Berg, David J. Kriegman
-
Patent number: 10318113Abstract: One or more embodiments of a digital content system allow a user to conveniently search and/or navigate through a collection of digital content items. In particular, a user can interact with a client device to search for and identify one or more digital content items within a collection of digital content items. For example, the digital content system may provide a photo from a collection of photos via a graphical user interface. The digital content system can receive a user input identifying a selection of one or more visual features within the photo. Based on the selected visual feature(s), the digital content system may identify photos from the collection of photos that include the identified visual feature(s) and provide access to the identified photos via the graphical user interface.Type: GrantFiled: July 11, 2016Date of Patent: June 11, 2019Assignee: DROPBOX, INC.Inventors: Peter N. Belhumeur, David J. Kriegman, Thomas Berg
-
Patent number: 10255527Abstract: The present disclosure is directed toward systems and methods that enable more accurate digital object classification. In particular, disclosed systems and methods address inaccuracies in digital object classification introduced by variations in classification scores. Specifically, in one or more embodiments, disclosed systems and methods generate probability functions utilizing digital test objects and transform classifications scores into normalized classification scores utilizing probability functions. Disclosed systems and methods utilize normalized classification scores to more accurately classify and identify digital objects in a variety of applications.Type: GrantFiled: September 21, 2017Date of Patent: April 9, 2019Assignee: DROPBOX, INC.Inventors: David J. Kriegman, Peter N. Belhumeur, Thomas Berg, Nils Peter Welinder
-
Publication number: 20180025251Abstract: The present disclosure is directed toward systems and methods to quickly and accurately identify boundaries of a displayed document in a live camera image feed, and provide a document boundary indicator within the live camera image feed. For example, systems and methods described herein utilize different display document detection processes in parallel to generate and provide a document boundary indicator that accurately corresponds with a displayed document within a live camera image feed. Thus, a user of the mobile computing device can easily see whether the document identification system has correctly identified the displayed document within the camera viewfinder feed.Type: ApplicationFiled: July 24, 2017Publication date: January 25, 2018Inventors: Nils Peter Welinder, Peter N. Belhumeur, Ying Xiong, Jongmin Baek, Simon Kozlov, Thomas Berg, David J. Kriegman
-
Publication number: 20180024974Abstract: The present disclosure is directed toward systems and methods that efficiently and effectively generate an enhanced document image of a displayed document in an image frame captured from a live image feed. For example, systems and methods described herein apply a document enhancement process to a displayed document in an image frame that result in an enhanced document image that is cropped, rectified, un-shadowed, and with dark text against a mostly white background. Additionally, systems and method described herein determine whether a stored digital content item includes a displayed document. In response to determining that a stored digital content item does include a displayed document, systems and methods described herein generate an enhanced document image of a displayed document included in the stored digital content item.Type: ApplicationFiled: July 24, 2017Publication date: January 25, 2018Inventors: Nils Peter Welinder, Peter N. Belhumeur, Ying Xiong, Jongmin Baek, Simon Kozlov, Thomas Berg, David J. Kriegman
-
Publication number: 20180012109Abstract: The present disclosure is directed toward systems and methods that enable more accurate digital object classification. In particular, disclosed systems and methods address inaccuracies in digital object classification introduced by variations in classification scores. Specifically, in one or more embodiments, disclosed systems and methods generate probability functions utilizing digital test objects and transform classifications scores into normalized classification scores utilizing probability functions. Disclosed systems and methods utilize normalized classification scores to more accurately classify and identify digital objects in a variety of applications.Type: ApplicationFiled: September 21, 2017Publication date: January 11, 2018Inventors: David J. Kriegman, Peter N. Belhumeur, Thomas Berg, Nils Peter Welinder
-
Patent number: 9846822Abstract: The present disclosure is directed toward systems and methods that enable more accurate digital object classification. In particular, disclosed systems and methods address inaccuracies in digital object classification introduced by variations in classification scores. Specifically, in one or more embodiments, disclosed systems and methods generate probability functions utilizing digital test objects and transform classifications scores into normalized classification scores utilizing probability functions. Disclosed systems and methods utilize normalized classification scores to more accurately classify and identify digital objects in a variety of applications.Type: GrantFiled: December 31, 2015Date of Patent: December 19, 2017Assignee: DROPBOX, INC.Inventors: David J. Kriegman, Peter N. Belhumeur, Thomas Berg, Nils Peter Welinder
-
Publication number: 20170193337Abstract: The present disclosure is directed toward systems and methods that enable more accurate digital object classification. In particular, disclosed systems and methods address inaccuracies in digital object classification introduced by variations in classification scores. Specifically, in one or more embodiments, disclosed systems and methods generate probability functions utilizing digital test objects and transform classifications scores into normalized classification scores utilizing probability functions. Disclosed systems and methods utilize normalized classification scores to more accurately classify and identify digital objects in a variety of applications.Type: ApplicationFiled: December 31, 2015Publication date: July 6, 2017Inventors: David J. Kriegman, Peter N. Belhumeur, Thomas Berg, Nils Peter Welinder
-
Publication number: 20160320932Abstract: One or more embodiments of a digital content system allow a user to conveniently search and/or navigate through a collection of digital content items. In particular, a user can interact with a client device to search for and identify one or more digital content items within a collection of digital content items. For example, the digital content system may provide a photo from a collection of photos via a graphical user interface. The digital content system can receive a user input identifying a selection of one or more visual features within the photo. Based on the selected visual feature(s), the digital content system may identify photos from the collection of photos that include the identified visual feature(s) and provide access to the identified photos via the graphical user interface.Type: ApplicationFiled: July 11, 2016Publication date: November 3, 2016Inventors: Peter N. Belhumeur, David J. Kriegman, Thomas Berg
-
Patent number: 9448704Abstract: One or more embodiments of a digital content system allow a user to conveniently search and/or navigate through a collection of digital content items. In particular, a user can interact with a client device to search for and identify one or more digital content items within a collection of digital content items. For example, the digital content system may provide a photo from a collection of photos via a graphical user interface. The digital content system can receive a user input identifying a selection of one or more visual features within the photo. Based on the selected visual feature(s), the digital content system may identify photos from the collection of photos that include the identified visual feature(s) and provide access to the identified photos via the graphical user interface.Type: GrantFiled: April 29, 2015Date of Patent: September 20, 2016Assignee: DROPBOX, INC.Inventors: Peter N. Belhumeur, David J. Kriegman, Thomas Berg
-
Patent number: 9275273Abstract: A system is provided for localizing parts of an object in an image by training local detectors using labeled image exemplars with fiducial points corresponding to parts within the image. Each local detector generates a detector score corresponding to the likelihood that a desired part is located at a given location within the image exemplar. A non-parametric global model of the locations of the fiducial points is generated for each of at least a portion of the image exemplars. An input image is analyzed using the trained local detectors, and a Bayesian objective function is derived for the input image from the non-parametric model and detector scores. The Bayesian objective function is optimized using a consensus of global models, and an output is generated with locations of the fiducial points labeled within the object in the image.Type: GrantFiled: July 7, 2014Date of Patent: March 1, 2016Inventors: Peter N. Belhumeur, David W. Jacobs, David J. Kriegman, Neeraj Kumar
-
Publication number: 20150078631Abstract: A system is provided for localizing parts of an object in an image by training local detectors using labeled image exemplars with fiducial points corresponding to parts within the image. Each local detector generates a detector score corresponding to the likelihood that a desired part is located at a given location within the image exemplar. A non-parametric global model of the locations of the fiducial points is generated for each of at least a portion of the image exemplars. An input image is analyzed using the trained local detectors, and a Bayesian objective function is derived for the input image from the non-parametric model and detector scores. The Bayesian objective function is optimized using a consensus of global models, and an output is generated with locations of the fiducial points labeled within the object in the image.Type: ApplicationFiled: July 7, 2014Publication date: March 19, 2015Inventors: Peter N. Belhumeur, David W. Jacobs, David J. Kriegman, Neeraj Kumar
-
Patent number: 8811726Abstract: A method is provided for localizing parts of an object in an image by training local detectors using labeled image exemplars with fiducial points corresponding to parts within the image. Each local detector generates a detector score corresponding to the likelihood that a desired part is located at a given location within the image exemplar. A non-parametric global model of the locations of the fiducial points is generated for each of at least a portion of the image exemplars. An input image is analyzed using the trained local detectors, and a Bayesian objective function is derived for the input image from the non-parametric model and detector scores. The Bayesian objective function is optimized using a consensus of global models, and an output is generated with locations of the fiducial points labeled within the object in the image.Type: GrantFiled: June 4, 2012Date of Patent: August 19, 2014Assignee: Kriegman-Belhumeur Vision Technologies, LLCInventors: Peter N. Belhumeur, David W. Jacobs, David J. Kriegman, Neeraj Kumar
-
Patent number: 8712189Abstract: Methods, systems, and media for swapping faces in images are provided. In some embodiments, a detected face and face data corresponding to an input image is received. A pose bin associated with the detected face is then identified based on the face data. Next, the detected face is aligned to a generic face associated with the pose bin. At least a portion of a candidate face associated with the pose bin is selected. The at least a portion of the candidate face is then copied to a copy of the input image that is aligned with the generic image to form a swapped-face image. The swapped-face image is next aligned to the input image to form an output image, and then the output image is outputted to a display.Type: GrantFiled: May 29, 2013Date of Patent: April 29, 2014Assignee: The Trustees of Columbia University in the City of New YorkInventors: Dimitri Bitouk, Neeraj Kumar, Peter N Belhumeur, Shree K Nayar
-
Publication number: 20130336600Abstract: Methods, systems, and media for swapping faces in images are provided. In some embodiments, a detected face and face data corresponding to an input image is received. A pose bin associated with the detected face is then identified based on the face data. Next, the detected face is aligned to a generic face associated with the pose bin. At least a portion of a candidate face associated with the pose bin is selected. The at least a portion of the candidate face is then copied to a copy of the input image that is aligned with the generic image to form a swapped-face image. The swapped-face image is next aligned to the input image to form an output image, and then the output image is outputted to a display.Type: ApplicationFiled: May 29, 2013Publication date: December 19, 2013Inventors: Dmitri Bitouk, Neeraj Kumar, Peter N. Belhumeur, Shree K. Nayar
-
Patent number: 8571332Abstract: Methods, systems, and media for automatically classifying face images are provided. In some embodiments, features of the face image to be classified for an attribute are selected, wherein each of the features corresponds to a different region of the face image and specifies one or more of a type of pixel data to be evaluated for the region, a normalization to be applied for the region, and an aggregation to be applied for the region. The face image is classified with respect to the attribute based on the features of the image, and the attribute and a confidence value are assigned to the face image based on the classifying. A query is received from a user, and the attribute is identified as corresponding to the query. The face image is determined as corresponding to the attribute, and the face image is identified to the user as corresponding to the query.Type: GrantFiled: March 19, 2009Date of Patent: October 29, 2013Assignee: The Trustees of Columbia University in the City of New YorkInventors: Neeraj Kumar, Peter N Belhumeur, Shree K Nayar, Alexander C Berg
-
Patent number: 8472722Abstract: Methods, systems, and media for swapping faces in images are provided. In some embodiments, a detected face and face data corresponding to an input image is received. A pose bin associated with the detected face is then identified based on the face data. Next, the detected face is aligned to a generic face associated with the pose bin. At least a portion of a candidate face associated with the pose bin is selected. The at least a portion of the candidate face is then copied to a copy of the input image that is aligned with the generic image to form a swapped-face image. The swapped-face image is next aligned to the input image to form an output image, and then the output image is outputted to a display.Type: GrantFiled: July 26, 2010Date of Patent: June 25, 2013Assignee: The Trustees of Columbia University in the City of New YorkInventors: Shree K. Nayar, Peter N. Belhumeur, Dimitri Bitouk, Neeraj Kumar
-
Publication number: 20120308124Abstract: A method is provided for localizing parts of an object in an image by training local detectors using labeled image exemplars with fiducial points corresponding to parts within the image. Each local detector generates a detector score corresponding to the likelihood that a desired part is located at a given location within the image exemplar. A non-parametric global model of the locations of the fiducial points is generated for each of at least a portion of the image exemplars. An input image is analyzed using the trained local detectors, and a Bayesian objective function is derived for the input image from the non-parametric model and detector scores. The Bayesian objective function is optimized using a consensus of global models, and an output is generated with locations of the fiducial points labeled within the object in the image.Type: ApplicationFiled: June 4, 2012Publication date: December 6, 2012Applicant: KRIEGMAN-BELHUMEUR VISION TECHNOLOGIES, LLCInventors: Peter N. Belhumeur, David W. Jacobs, David J. Kriegman, Neeraj Kumar
-
Publication number: 20110243461Abstract: Methods, systems, and media for automatically classifying face images are provided. In some embodiments, features of the face image to be classified for an attribute are selected, wherein each of the features corresponds to a different region of the face image and specifies one or more of a type of pixel data to be evaluated for the region, a normalization to be applied for the region, and an aggregation to be applied for the region. The face image is classified with respect to the attribute based on the features of the image, and the attribute and a confidence value are assigned to the face image based on the classifying. A query is received from a user, and the attribute is identified as corresponding to the query. The face image is determined as corresponding to the attribute, and the face image is identified to the user as corresponding to the query.Type: ApplicationFiled: March 19, 2009Publication date: October 6, 2011Inventors: Shree K Nayar, Peter N Belhumeur, Neeraj Kumar, Alexander C Berg