Jayant Kumar

US Patent:

20130121610, May 16, 2013

Filed:

Nov 10, 2011

Appl. No.:

13/293845

Inventors:

Francine CHEN - Menlo Park CA, US
Jayant KUMAR - College Park MD, US

Assignee:

FUJI XEROX CO., LTD. - Tokyo

International Classification:

G06K 9/40

US Classification:

382266

Abstract:

Systems and methods described herein are directed to estimating the sharpness of images or photos such as document pages or scenes. The systems and methods are based on calculating a difference in grayscale values between pixels at an edge of an image. The differences may be accomplished by taking a slope of grayscale values between two pixels at an edge window, and estimating the edge sharpness based on the slope, multiple slopes, or differences in the slopes.

US Patent:

20140050367, Feb 20, 2014

Filed:

Aug 15, 2012

Appl. No.:

13/586784

Inventors:

Francine CHEN - Menlo Park CA, US
Scott CARTER - Los Altos CA, US
Laurent DENOUE - Verona, IT
Jayant KUMAR - Hyattsville MD, US

Assignee:

FUJI XEROX CO., LTD. - Tokyo

International Classification:

G06K 9/00

US Classification:

382112

Abstract:

A method for providing real-time feedback of an estimated quality of a captured final image, the method including obtaining a preliminary image, calculating a quality score of the preliminary image, and in response to the quality score of the preliminary image exceeding a threshold quality value, taking a first action.

US Patent:

20220253435, Aug 11, 2022

Filed:

Feb 10, 2021

Appl. No.:

17/172986

Inventors:

- SAN JOSE CA, US
Venkat Barakam - Cupertino CA, US
Benjamin Leviant - Mountain View CA, US
Amine Ben Khalifa - Sunnyvale CA, US
Kerem Turgutlu - Ames IA, US
Jayant Kumar - San Jose CA, US
Sumeet Zaverilal Gala - Milpitas CA, US
Gaurav Kukal - Fremont CA, US
Vipul Dalal - Cupertino CA, US

International Classification:

G06F 16/245
G06N 3/04
G06N 3/08

Abstract:

Systems and methods for information retrieval are described. Embodiments generate a dense embedding for each of a plurality of media objects to be searched, generate a sparse embedding for each of the media objects using an encoder that takes the dense embedding as an input, wherein the sparse embedding satisfies a sparsity constraint that is applied to at least one layer of the encoder during training, and perform a search on the plurality of media objects based at least in part on the sparse embedding.

US Patent:

20210272253, Sep 2, 2021

Filed:

Feb 27, 2020

Appl. No.:

16/803332

Inventors:

- San Jose CA, US
Vipul Dalal - Cupertino CA, US
Vera Lychagina - San Jose CA, US
Shabnam Ghadar - Menlo Park CA, US
Rohith mohan Dodle - Fremont CA, US
Mina Doroudi - San Francisco CA, US
Midhun Harikumar - Santa Clara CA, US
Kannan Iyer - San Ramon CA, US
Jayant Kumar - San Jose CA, US
Gaurav Kukal - Fremont CA, US
Daniel Miranda - San Jose CA, US
Charles R. McKinney - Aptos CA, US
Archit Kalra - Fremont CA, US

International Classification:

G06T 5/50
G06K 9/00
G06T 7/11
G06K 9/62
G06T 7/70

Abstract:

The present disclosure relates to an image merging system that automatically and seamlessly detects and merges missing people for a set of digital images into a composite group photo. For instance, the image merging system utilizes a number of models and operations to automatically analyze multiple digital images to identify a missing person from a base image, segment the missing person from the second image, and generate a composite group photo by merging the segmented image of the missing person into the base image. In this manner, the image merging system automatically creates merged group photos that appear natural and realistic.

US Patent:

20210248748, Aug 12, 2021

Filed:

Feb 12, 2020

Appl. No.:

16/789088

Inventors:

- San Jose CA, US
Jayant Kumar - San Jose CA, US
Jianming Zhang - Campbell CA, US
Zhe Lin - Fremont CA, US

Assignee:

Adobe Inc. - San Jose CA

International Classification:

G06T 7/11
G06T 3/40
G06T 7/194
G06N 3/04

Abstract:

Techniques are disclosed for parsing a source image, to identify segments of one or more objects within the source image. The parsing is carried out by an image parsing pipeline that includes three distinct stages comprising three respectively neural network models. The source image can include one or more objects. A first neural network model of the pipeline identifies a section of the source image that includes the object comprising a plurality of segments. A second neural network model of the pipeline generates, from the section of the source image, a mask image, where the mask image identifys one or more segments of the object. A third neural network model of the pipeline further refines the identification of the segments in the mask image, to generate a parsed image. The parsed image identifies the segments of the object, by assigning corresponding unique labels to pixels of different segments of the object.