1988results about "Still image data clustering/classification" patented technology

The present disclosure relates to systems and methods for identifying products and monitoring planogram compliance using analysis of image data. In one implementation, the method may include accessing at least one planogram describing a desired placement of a plurality of product types on shelves of a plurality of retail stores; receiving image data from the plurality of retail stores; analyzing the image data to determine an actual placement of the plurality of product types on the shelves of the plurality of retail stores; determining at least one characteristic of planogram compliance based on detected differences between the at least one planogram and the actual placement of the plurality of product types on the shelves of the plurality of retail stores; and receiving checkout data from the plurality of retail stores reflecting sales of at least one product type from the plurality of product types.

A person-based digital photo clustering method, medium, and apparatus and a person-based digital photo albuming method, medium, and apparatus. The clustering method may include detecting a face area from photo data, extracting predetermined content-based face feature value information from the detected face area, obtaining person-based photo clustering hint information, and performing person-based clustering by applying photo clustering hint information to the extracted content-based face feature value information. The person-based digital photo albuming method may include generating photo description information for identifying at least a photo, generating albuming tool description information supporting person-based photo clustering and including at least a parameter for person-based photo clustering, performing photo albuming including person-based clustering by using at least photo description information and albuming tool description information, generating predetermined photo group description information from photo albumed data, and generating predetermined photo album information by using photo description information and photo group description information.

The present invention relates to managing media items in data processing terminals. More particularly, the present invention is directed to a method, a device and a computer program product for arranging, viewing and querying media items organized in hierarchical multidimensional clusters in mobile terminals. Media items are arranged by clustering with multiple dimensions, wherein they are queried by defining the first entry for one dimension, wherein the next entry is based on the other dimension from the media items fulfilling the first entry.

A method and system for categorizing non-textual subject data, such as digital images, content-based data and meta-data to determine outcomes of classification tasks. The meta-data is indicative of the operational conditions of a recording device during the capturing of the content-based data. For example, the non-textual subject data may be a digital image captured by a digital camera, and the meta-data may include automatic gain setting, film speed, shutter speed, aperture/lens index, focusing distance, date and time, and flash/no flash operation. The subject image is tagged with selected classifiers by subjecting the image to a series of classification tasks utilizing both content-based data and meta data to determine classifiers associated with the subject image.

A digital picture frame including a camera integrated with the frame, and a network connection module allowing the frame for direct contact and upload of photos from electronic devices or from a user's social media account or her or his community members' social media accounts. The integrated camera is used to automatically determine an identity of a frame viewer, and a viewer profile automatically determined from the identity of the viewer. The displayed photos are automatically shown and/or changed according to the detected viewers.

A system for processing images captured in a retail store is provided. The system may include a processor configured to: access a database storing a group of product models; receive an image depicting at least part of a store shelf having a plurality of products of a same type displayed thereon; analyze the image and determine a first candidate type of the products based on the group of product models and the image analysis; determine a first confidence level associated with the first candidate type; when the first confidence level is below a confidence threshold, determine a second candidate type of the products using contextual information; determine a second confidence level associated with the determined second candidate type of the plurality of products; and when the second confidence level is above the confidence threshold, initiate an action to update the group of product models stored in the database.

In an image classification method, a plurality of grouping values are received. The grouping values each have an associated image. An average of the grouping values is calculated. A variance metric of the grouping values, relative to the average is computed. A grouping threshold is determined from the variance metric. Grouping values beyond the grouping threshold are identified as group boundaries. The images are assigned to a plurality of groups based upon the group boundaries.

Various aspects of the subject technology relate to systems, methods, and machine-readable media for clustering search results based on image composition. A system may, for each image in a set of images, determine a compositional vector representing one or more objects and corresponding locations within the image using a trained computer-operated convolutional neural network. The system may provide each image through a clustering algorithm to produce one or more clusters based on compositional similarity. The system may provide images from the set of images clustered by composition, in which the images include a different listing of images for each of the one or more clusters. The system may provide a prioritized listing of images responsive to a user search query, in which the prioritized listing of images includes a different listing of images for each cluster of compositional similarity based on the metadata of each image associated with the cluster.

A system for acquiring images of products in a retail store is disclosed. The system may include a first housing configured for location on a retail shelving unit, and an image capture device included in the first housing and configured relative to the first housing such that an optical axis of the image capture device is directed toward an opposing retail shelving unit when the first housing is fixedly mounted on the retail shelving unit. The system may further include a second housing configured for location on the retail shelving unit separate from the first housing, the second housing containing a processor configured to control the image capture device and a network interface for communicating with a remote server. The system may further include a data conduit configured to enable transfer of control signals from the processor to the image capture device.

Process-response statistical modeling of visual images can be used in determining similarity between visual images. Evaluation of the content of visual images—and, in particular, image similarity determinations—can be used in effecting a variety of interactions (e.g., searching, indexing, grouping, summarizing, annotating, keyframing) with a collection of visual images.

The embodiment of the invention discloses a vehicle loss assessment image obtaining method and device, a server, and terminal equipment. A client obtains photographed video data and transmits the photographed video data to the server. The client receives the information of a specified damaged part of a damaged vehicle, and transmits the information of the damaged part to the server. The server receives the photographed video data and the information of the damaged part which are uploaded by the client, and extracts a video image in the photographed video data, and classifies the video image according to the information of the damaged part, and determines a candidate image classification set of the damaged part. The server selects a loss assessment image of the vehicle from the candidate image classification set according to a preset screening condition. According to all embodiment of the invention, the method can achieve the automatic and quick generation of a high-quality loss assessment image meeting the loss assessment processing demands, meets the loss assessment processing demands, and improves the obtaining efficiency of a loss assessment image.

The invention discloses a neural network construction method and device and an image processing method and device in the field of computer vision in the artificial intelligence field. The neural network construction method comprises the following steps: determining a search space and a plurality of construction units; stacking the plurality of construction units to obtain a search network, with the search network being a neural network for searching a neural network structure; and optimizing a network structure of a construction unit in the search network in the search space, wherein in the optimization process, the search space is gradually reduced, the number of the construction units is gradually increased, and due to the reduction of the search space and the increase of the number of the construction units, the video memory consumption generated in the optimization process is within a preset range; and establishing the target neural network according to the optimized construction unit. Under the condition that video memory resources are certain, the neural network meeting application requirements well can be constructed.

A classification apparatus is provided. The classification apparatus includes a reader apparatus configured to receive visual information and textual information associated with the visual information and detect query relevant categorization information regarding products or service of interest to a user from the visual information and textual information associated with the visual information, a localize and identify apparatus configured to receive the visual information and the query relevant categorization information and selectively reduce the visual information to a relevant visual representation and detect further categorization information based on the relevant visual representation, and a deep learning processor apparatus comprising a unit classifier and a type classifier, wherein the unit classifier correlates the relevant visual representation with a specific product identification code, and the type classifier correlates the relevant visual representation with a type considered represented in the relevant visual representation.

The invention relates to the technical field of image processing, in particular to an image classification method and device, a storage medium and terminal equipment. The image classification method provided by the invention comprises the following steps: obtaining a to-be-classified image; inputting the to-be-classified image into a trained image classification model to obtain an image label output by the image classification model, wherein the image classification model comprises a teacher model and a student model, and the teacher model is a convolutional neural network model obtained by carrying out weak supervision training by utilizing noise data and non-noise data, and the student model is a model obtained by carrying out joint training according to a knowledge migration mechanism and a multi-task learning mechanism based on the teacher model; obtaining a classification result corresponding to the to-be-classified image according to the image label, using the teacher model to perform knowledge migration, and using the noise data and the non-noise data to perform multi-task learning training, thereby improving the classification efficiency and the classification accuracy of the image classification model.

The present disclosure generally relates to navigating a collection of media items. In accordance with one embodiment, a device displays a plurality of content items in a first layout, including a first content item at a first aspect ratio and a first size, a second content item, and a third content item. While displaying the plurality of content items in the first layout, the device detects a user input that includes a gesture, wherein the user input corresponds to a request to change a size of the first content item. In response to detecting the user input, and as the gesture progresses, the device changes the size of the first content item from the first size to a second size while concurrently gradually changing an aspect ratio of the first content item from the first aspect ratio to a second aspect ratio.

A digital image is automatically combined with text data by automatically classifying the image according to a predetermined set of categories, and automatically selecting text data from a repository, wherein the selected text data match the category of the image according to a predetermined criterion.

A method and apparatus for situation-based clustering digital photos, and a digital photo albuming system and method using the same. A situation-based digital photo clustering method of clustering digital photos based on a situation when a photo is taken includes: extracting photographing data information including at least a photographing time feature value from a digital photo file and extracting a content-based feature value from contents of a digital photo of the digital photo file; assigning an importance degree to each extracted photographing time feature value and content-based feature value and combining the values; and hierarchically clustering photographing situations using feature value information, the feature value information being the extracted photographing time feature value and content-based feature value combined with respect to the assigned degrees of importance.

An information processing apparatus includes an acquisition unit for acquiring a plurality of pieces of information to be grouped, the pieces of information including time information as attribute information, and an arithmetic unit for grouping the pieces of information acquired by the acquisition unit so that a portion in which a level of variation of intervals among times represented by the time information included in the pieces of information is greater than at least one threshold value is used as a boundary between groups.

The embodiment of the invention discloses an image-text matching method and device, a storage medium and equipment, and belongs to the technical field of computers. The image-text matching method comprises the steps of obtaining an image and a text to be matched; generating a candidate instance feature set according to the image; aggregating candidate instance features in the candidate instance feature set by using a self-attention mechanism to obtain an instance feature set, each instance feature in the instance feature set corresponding to one object in the image; encoding the text to obtaina text vector; and calculating a matching degree between the image and the text according to the instance feature set and the text vector. According to the embodiment of the invention, the realization difficulty of image-text matching can be simplified, and the accuracy of image-text matching can be improved.

The invention discloses a method for realizing gathering and scattering visualization of mass surface feature mark points through layered grid division. The method comprises: acquiring the current zoom ratio of a map, a preset gathering and scattering hierarchy calculation rule is combined to determine whether gathering display or scattering display is currently carried out, and the scattering display execution step comprises the steps of obtaining a corresponding longitude and latitude range on the map according to pixel coordinates of a visual display area; according to the latitude and longitude range, inquiring the dynamic index to obtain marker information in the latitude and longitude range; screening the marker information by using a grid filtering algorithm according to a backwardfirst-in first-out principle; and rendering the screened marker information to a map by adopting a coordinate transformation algorithm for scattered display. According to the method, the display efficiency of mass data on the map can be improved, the identification degree of scattered display of mass points is improved, and map interactive query is achieved.

A method includes receiving, at a mobile device, a selection of a first media item transferred from the mobile device to a remote device. The method also includes selecting, by the mobile device, a second media item to transfer to the remote device based on a similarity of the second media item to the first media item. The method further includes generating, by the mobile device, packaged moment data to transfer to the remote device, where the packaged moment data includes media data associated with multiple media items including the first media item and the second media item.

A property vector representing extractable measurable properties, such as musical properties, of a file is mapped to semantic properties for the file. This is achieved by using artificial neural networks “ANNs” in which weights and biases are trained to align a distance dissimilarity measure in property space for pairwise comparative files back towards a corresponding semantic distance dissimilarity measure in semantic space for those same files. The result is that, once optimised, the ANNs can process any file, parsed with those properties, to identify other files sharing common traits reflective of emotional perception, thereby rendering a more liable and true-to-life result of similarity/dissimilarity. This contrasts with simply training a neural network to consider extractable measurable properties that, in isolation, do not provide a reliable contextual relationship into the real-world.

A system and method of saving digital content classified by person-based clustering. The system for saving digital content classified by person-based clustering, includes a database to save a plurality of digital content classified by person-based clustering; a data structure generation unit to generate a data structure composed of a plurality of nodes using the plurality of digital content; a face recognition unit to extract a face descriptor of new digital content to be saved in the database; a cluster classification unit to classify the new digital content and the plurality of digital content by the person-based clustering using the extracted face descriptor, and a data structure update unit to update the data structure according to the classification.

The invention provides a data classification method and device, a computer device and a storage medium. The method comprises the steps of obtaining a to-be-labeled data set; labeling the data set through a labeling function to obtain an initial label of the data set; calculating a pair correlation of the identification function according to the initial label, and constructing a generation model ofthe marking function according to the pair correlation; estimating a probability label of the data set according to the generation model; training a discrimination model according to the probabilitylabel to obtain a trained discrimination model; and inputting to-be-classified data into the trained discrimination model to obtain the category of the to-be-classified data. According to the method,the marking efficiency and accuracy of the training data are improved, the training data can be utilized to quickly train the discrimination model, and the discrimination model is utilized to realizethe quick and accurate data classification.

The invention discloses a small sample learning algorithm based on covariance measurement, and belongs to the field of computer vision. The algorithm comprises the steps: (1) introducing an interpolation training mechanism to learn migratable knowledge; (2) designing a local covariance representation, and then embedding the local covariance representation into a deep network to learn and express each concept; (3) constructing a covariance measurement layer to measure the distribution consistency between a query sample and the concepts based on the local covariance representation. According tothe method, a novel and concise end-to-end covariance measurement network CovaMNet is provided, a second-order local covariance representation is designed to replace traditional first-order concept representation, and a new covariance measurement function is provided. A comparison experiment result on a plurality of reference data sets is analyzed to obtain that the CovaMNet framework provided bythe invention shows a competitive effect on a general small sample classification task and a fine-grained small sample classification task.