Google Inc.

Fujifilm Corporation

Thomas Leung

Yang Song

Troy Chinen

Jay Yagnik

Charles Joseph Rosenberg

John Zhang

Shumeet Baluja

Sergey Ioffe

Yushi Jing

A classifier training system trains unified classifiers for categorizing videos representing different categories of a category graph. The unified classifiers unify the outputs of a number of separate initial classifiers trained from disparate subsets of a training set of media items. The training process takes into account the relationships that exist between the various categories of the category graph by relating scores associated with related categories, thus enhancing the accuracy of the unified classifiers.

Cluster-based video classification

An image quality subsystem computes quality scores for images that represent a measure of visual quality of the images. Initial quality scores and query specific quality scores can be computed for the images based on image feature values for the images and a transformation factor that represents a measure of importance of image quality for computing relevance scores for images. The initial quality scores are query independent quality scores that are computed for the images and can be used as a factor for computing relevance scores for the image relative to any query. Query specific quality scores are computed for images that are identified as relevant for a particular query based on the initial quality scores and a query specific transformation factor for the particular query. Adjusted relevance scores for the images can be computed based on the initial quality scores or the query specific quality scores.

Image selection based on image quality

Methods, systems, and apparatus, including computer programs encoded on computer storage media, for clustering images. In one aspect a system includes one or more computers configured to, for each of a plurality of digital images, associate extrinsic image-related information with each individual image, the extrinsic image-related information including text information and co-click data for the individual image, assign images from the plurality of images to one or more of the clusters of images based on the extrinsic information associated with each of the plurality of images, receive in the search system a user query from a user device, identify by operation of the search system one or more clusters of images that match the query, and provide one or more cluster results, where each cluster result provides information about an identified cluster.

Clustering images

A classifier training system trains unified classifiers for categorizing videos representing different categories of a category graph. The unified classifiers unify the outputs of a number of separate initial classifiers trained from disparate subsets of a training set of media items. The training process divides the training set into a number of bags, and applies a boosting algorithm to the bags, thus enhancing the accuracy of the unified classifiers.

Multiple-instance-learning-based video classification

Image quality measures

A video hosting service automatically identifies, in a video database, a set of videos associated with an advertiser, and presents the identified videos to the advertiser for consideration. The videos may be selected based on analysis of their video content for images of logos associated with the advertisers. The video hosting service may then receive from the advertisers a listing of which of the presented videos should be given an award.

Rewarding creative use of product placements in user-contributed videos

A method and an apparatus process digital images. The method according to one embodiment accesses element data representing a plurality of elements belonging to a plurality of digital images; performs a similarity analysis between the elements from the plurality of elements to obtain inter-relational data results relating to the elements; and performs clustering of the plurality of elements, the step of performing clustering including incorporating in the inter-relational data results at least one hard constraint relating to elements from the plurality of elements, to obtain constrained inter-relational data results, performing a spectral analysis to obtain eigenvector results from the constrained inter-relational data results, and performing discretization of the eigenvector results using constrained clustering with a criterion to enforce the at least one hard constraint to obtain clusters.

Method and apparatus for performing constrained spectral clustering of digital image data

A method and an apparatus process digital images. The method according to one embodiment accesses digital data representing a plurality of digital images including a plurality of persons; performs face recognition to determine first scores relating to similarity between faces of the plurality of persons; performs clothes recognition to determine second scores relating to similarity between clothes of the plurality of persons; provides a plurality of formulas for estimating a probability of a face from the faces and a clothes from the clothes to belong to a person from the plurality of persons, wherein at least one formula of the plurality of formulas utilizes a first score and a second score, and at least one formula of the plurality of formulas utilizes only one score of a first score and a second score; and selects a formula from the plurality of formulas based on availability of a first score from the first scores for two persons from the plurality of persons, and availability of a second score from the second scores for the two persons, the selected formula estimating a probability relating to similarity of identities of the two persons.

Method and apparatus for adaptive context-aided human classification

An exemplary illumination normalization method is provided which includes receiving an input image having at least one spurious edge directly resulting from illumination, performing anisotropic diffusion on the input image to form a diffusion image, and removing at the least one spurious edge using the diffusion image. Another embodiment consistent with the invention is an apparatus for performing illumination normalization in an image which includes a processor operably coupled to a memory storing input image data which contains an object of interest having at least one spurious edge directly resulting from illumination, a model of a representative object of interest, and functional processing units for controlling image processing, wherein the functional processing units further include a model based anisotropic diffusion module which predicts edge information regarding the object of interest based upon the model, and produces a reflectance estimation utilizing the predicted edge information.

Method and apparatus for diffusion based illumination normalization

A sequence of images depicts a foreground object in motion. A base image is selected, and the other images in the sequence are co-registered with the base image in order to align the images to a common coordinate system. A background image and a binary foreground mask are generated from the sequence of aligned images. By applying the foreground mask to a chosen one of the aligned images, a representation of the moving object is extracted. After blurring the background image, the extracted representation may be superimposed onto the blurred background image to produce a new image.

Segmenting images and simulating motion blur using an image sequence

A method for image relighting is presented which receives an input image having at least one spurious edge directly resulting from a first illumination present when the input image was acquired, performs anisotropic diffusion on the input image to form a diffusion image, removes the first illumination using the diffusion image to generate a reflectance image and applies a second illumination to the reflectance image. An apparatus for relighting is presented which includes a processor operably coupled to memory storing input image having a first illumination present when the input image was acquired, and functional processing including an anisotropic diffusion module to perform anisotropic diffusion on the input image to form a diffusion image, a combination module which removes the first illumination using the diffusion image to generate a reflectance image, a second illumination module which generates a second illumination, and a lighting application model which applies the second illumination.

Growing community of inventors

Santa Clara, CA, United States of America

Thomas Leung