Abstract: A method is provided for controlling one or more lighting devices to generate a dynamic light effect based on an image. A first area in the image is selected based on first selection criteria (e.g. object recognition). Within this first area, a sequence of sub-areas (260, 262, 264) are selected based on second selection criteria (e.g. zooming in to a recognized object). Color values are extracted from each sub-area and the one or more lighting devices are controlled based on the extracted color values.

Abstract: The present disclosure relates to an image re-ranking method, which includes: performing image searching by using an initial keyword, obtaining, by calculation, an anchor concept set of a search result according to the search result corresponding to the initial keyword, obtaining, by calculation, a weight of a correlation between anchor concepts in the anchor concept set, and forming an anchor concept graph ACG by using the anchor concepts in the anchor concept set as vertexes and the weight of the correlation between anchor concepts as a weight of a side between the vertexes; acquiring a positive training sample by using the anchor concepts, and training a classifier by using the positive training sample; obtaining a concept projection vector by using the ACG and the classifier; calculating an ACG distance between images in the search result corresponding to the initial keyword; and ranking the images according to the ACG distance.

Abstract: Systems and methods are disclosed for estimating aesthetic quality of digital images using deep learning. In particular, the disclosed systems and methods describe training a neural network to generate an aesthetic quality score digital images. In particular, the neural network includes a training structure that compares relative rankings of pairs of training images to accurately predict a relative ranking of a digital image. Additionally, in training the neural network, an image rating system can utilize content-aware and user-aware sampling techniques to identify pairs of training images that have similar content and/or that have been rated by the same or different users. Using content-aware and user-aware sampling techniques, the neural network can be trained to accurately predict aesthetic quality ratings that reflect subjective opinions of most users as well as provide aesthetic scores for digital images that represent the wide spectrum of aesthetic preferences of various users.

Abstract: Introduced here is a machine learning related technique for supplying an observed model additional training data based upon previously received training data. To determine textual content of a character string based on a digital image that includes a handwritten version of the character string a substantial amount of training data is used. The character string can include one or more characters, and the characters can include any of letters, numerals, punctuation marks, symbols, spaces, etc. Disclosed herein is a technique to determine variations between different images of matching known character strings and substitute those variations into the images in order to create more images with the same known character string.

Abstract: A method that receives a visual media item and determines an identity of at least one intended viewer of the visual media item is disclosed. The method may further identify a visual representation of an object that is comprised by the visual media item and determine a relevance of the object based, at least in part, on the identity. The method may further generate a modified visual media item such that the modified visual media item differs from the visual media item, at least, by visual emphasis of the visual representation of the object.

Abstract: A lip-makeup analysis method adopted by a body information analysis apparatus includes following steps: obtaining an external image through an image capturing module of the body information analysis apparatus; performing positioning actions to each part of a face once the face is recognized from the external image; obtaining a position of the lips after completing the positioning actions; analyzing an average color value of lip-makeup of the lips; comparing the average color value of lip-makeup with a default color value; displaying a comparison result by a display module of the body information analysis apparatus; and, re-executing above steps before assisting actions are finished.

Abstract: A real-world object is recognized from image data based on derived image characteristics, which are used to derive search information to conduct a search. The search returns object information which is used to execute a software process by a mobile device.

Abstract: Systems and methods for digital video stabilization via constraint-based rotation smoothing are provided. Digital video data including a set of image frames having associated time stamps and a set of camera orientation data having associated time stamps may be provided. A smoothed set of camera orientation data may be generated by minimizing a rate of rotation between successive image frames while minimizing an amount of empty regions in a resulting set of smoothed image frames reoriented based on the smoothed set of camera orientation data.

Abstract: Embodiments of the present invention provide a system for transmitting data to networked resources based on extracted contextual information from said networked resources. The system gathers contextual data from devices on the network, extracts contextual information from the contextual data, and determines supplemental resources to provide to the networked devices based on the contextual information. The system may further provide an augmented reality environment to a networked device that integrates identifies objects in an image and integrates notifications of supplemental resources in the augmented reality environment.

Abstract: A mobile system that performs object recognition, identifying real-world objects in a scene from captured digital data of the scene. The mobile system identifies the real-world objects based on derived salient characteristics from the digital data, and retrieves stored object information corresponding to the identified object such that a CPU can execute a software process based on the retrieved object information about the real-world object.

Abstract: Embodiments are disclosed for detecting duplicate and near duplicate images. An exemplary method includes receiving an original image, preparing the image for fingerprinting, and calculating an image fingerprint, the fingerprint expressed as a sequence of numbers. The method further includes comparing the image fingerprint thus obtained with a set of previously stored fingerprints obtained from a set of previously stored images, and determining if the original image is either a duplicate or a near duplicate of an image in the set if the dissimilarity between the two fingerprints is less than a defined threshold T. Once a duplicate or near duplicate is detected, various defined actions may be taken, including culling the less desirable image or referring the redundancy to a user.

Abstract: A digital image of the object is captured and the object is recognized from plurality of objects in a database as part of a computer-based game. An information address corresponding to the object is then used to access content information associated with the identified object and interact with the game based on the content information.

Abstract: Methods, systems, and apparatus for receiving a query image, receiving one or more entities that are associated with the query image, identifying, for one or more of the entities, one or more candidate search queries that are pre-associated with the one or more entities, generating a respective relevance score for each of the candidate search queries, selecting, as a representative search query for the query image, a particular candidate search query based at least on the generated respective relevance scores and providing the representative search query for output in response to receiving the query image.

Abstract: Systems and methods are disclosed for providing purchase recommendations. According to some examples, a method may include: determining respective frequency distributions of a plurality of vehicle attributes, the frequency distributions being determined based on occurrences of values of the plurality of vehicle attributes in a set of vehicles; selecting a vehicle attribute from the plurality of vehicle attributes based on the frequency distributions of the plurality of vehicle attributes; transmitting, to a user device, an inquiry for user preference regarding the selected vehicle attribute; receiving, from the user device, a response indicating the user preference; and presenting, to the user device, a recommendation of one or more vehicles determined based on the received response.

Abstract: A distance measuring apparatus includes: an acquisition unit that acquires a first image at a first viewpoint where an object is irradiated with a first light including a pattern, a second image at a second viewpoint different from the first viewpoint where the object is irradiated with the first light, a third image at the first viewpoint where the object is irradiated with a second light not including a pattern, and a fourth image at the second viewpoint where the object is irradiated with the second light; and a control unit that acquires information corresponding to a distance, by employing a fifth image obtained based on a ratio the first image and the third image and a sixth image obtained based on a ratio of the second image and the fourth image.

Abstract: An optical tracker system is provided for tracking a location and orientation of a first object. The system includes a prediction module configured to determine a predicted location and orientation of the first object as a spatial prediction; a display module configured to generate display commands for a fiducial symbol based on the spatial prediction; a display unit configured to display the fiducial symbol based on the display commands; a camera unit configured to capture an image of the fiducial symbol displayed on the display device as a captured appearance; and a position determination module coupled to the camera unit and configured to receive the image from camera unit. The position determination module is configured to identify the fiducial symbol and determine the location and the orientation of the first object based on the captured appearance of the fiducial symbol.

Abstract: A computer-implemented method includes: monitoring, by a computing device, video or image data captured by a plurality of cameras implemented within a vehicle; monitoring, by the computing device, vehicle information associated with the vehicle, wherein the vehicle information includes sensor information reported by tire pressure sensors or impact sensors of the vehicle, or vehicle diagnostic information; detecting, by the computing device, satisfaction of particular video upload event criteria for which video should be uploaded for secure storage and future analysis, wherein the detecting the satisfaction of the video upload criteria is based on the monitoring the video or image data and the vehicle information; and uploading, by the computing device, video captured by a subset of the plurality of cameras to a video storage server, wherein the subset of the plurality of cameras include cameras capture the video or image data used to detect the satisfaction of the video upload event criteria or are associated

Abstract: A picture of a person's face is present in a mixed reality system. The mixed reality system has a monitoring or updating process that attempts to detect the presence of faces in the mixed reality. When a picture of a face is detected, the system detects edges of a picture frame or physical borders of the picture. If the picture is bounded by edges or is otherwise in a suitable physical form, then a canvas object is generated. The canvas object is arranged in the mixed reality to coincide with the picture of the face. A face recognition algorithm is used to find an identity of the face. Information updates specific to the identity of the face are obtained, applied to the canvas, and the canvas is rendered in the mixed reality. A viewer of the mixed reality will observe the information updates replacing or overlaying the picture.

Abstract: Text region detection techniques and systems for digital images using image tag filtering are described. These techniques and systems support numerous advantages over conventional techniques through use of image tags to filter text region candidates. A computing device, for instance, may first generate text region candidates through use of a variety of different techniques, such as text line detection. The computing device then assigns image tags to the text region candidates. The assigned image tags are then used by the computing device to filter the text region candidates based on whether image tags assigned to respective candidates are indicative of text.

Abstract: According to one embodiment, a strapping apparatus includes a setting device, a determination device, a strap band supply device, an acquisition device, a color decision device, and a warning presentation device. The setting device sets a denomination of paper sheets to be strapped. The determination device determines a color of a strap band to strap a bundle of the paper sheets, based on the denomination. The strap band supply device supplies strap bands from a first strap band reel previously stored. The acquisition device acquires an image of the first strap band reel. The color decision device decides a color of the first strap band reel, based on the image. The warning presentation device presents a warning if the color of the first strap band reel is different from the color determined by the determination device.

Abstract: An example radial histogram matching system may generate a target radial histogram by identifying pixels in an input digital image that are determined to be on, traversing each of the pixels in the input digital image one time to determine how many of the pixels that are turned on are contained in each sector of a circle and assigning to each elements of the target radial histogram the number of on pixels in that element. The system may also compare the target radial histogram to each of an initial and a sequence of rotated radial histograms to determine a match score for each of the comparisons, and identify an offset rotation between a baseline pattern digital image and the input digital image based on the match scores.

Abstract: Systems and methods for providing convolutional neural network based image synthesis using localized loss functions is disclosed. A first image including desired content and a second image including a desired style are received. The images are analyzed to determine a local loss function. The first and second images are merged using the local loss function to generate an image that includes the desired content presented in the desired style. Similar processes can also be utilized to generate image hybrids and to perform on-model texture synthesis. In a number of embodiments, Condensed Feature Extraction Networks are also generated using a convolutional neural network previously trained to perform image classification, where the Condensed Feature Extraction Networks approximates intermediate neural activations of the convolutional neural network utilized during training.

Abstract: A probabilistic model is provided based on an output of a matching procedure that matches a particular object to representations of objects, where the probabilistic model relates a probability of an object being present to a number of matching features. The probabilistic model is used for detecting whether a particular object is present in received visual data.

Abstract: A method and apparatus for generating image information. A specific embodiment of the method includes: acquiring a rotation angle and a scaling of a target image relative to a reference image, and acquiring an intrinsic parameter of a video camera collecting the reference image; generating a projective transformation matrix based on the rotation angle, the scaling and the intrinsic parameter; generating a coordinate offset of the each neighborhood pixel of the feature point in the reference image; generating, based on a coordinate of the feature point in the reference image, the coordinate offset of the each neighborhood pixel of the feature point in the reference image and the projective transformation matrix, coordinates of the each neighborhood pixel of the feature point in the reference image; and generating a feature point descriptor of the feature point.

Abstract: A platform, device and process for capturing images of the surface of a gaming table and determining the quantity, identity, and arrangement of chips bet at a gaming table. Image data is captured corresponding to the one or more chips positioned in at least one betting area on a gaming surface of the respective gaming table and the data is processed to filter out the background, establish a two dimensional grid of points of interests and corresponding histograms for classifying the one or more chips through identifying a dominant classification of each row in the grid of points of interests.

Abstract: An image forming apparatus includes an image forming unit, an image reading unit, an image inspecting unit and a history generating unit. The image forming unit forms an image on a sheet. The image reading unit reads a sheet face on which the image is formed, thereby generating a scan image. The image inspecting unit compares the generated scan image with a normal image to detect an error in the scan image. When the image inspecting unit detects the error, the history generating unit generates detection information on the detected error and embeds the generated detection information in the scan image, thereby generating a history image.

Abstract: A system, computer program product, and computer-executable method of managing one or more tiers of memory of a host computing system, the system, computer program product, and computer-executable method including accessing a portion of raw data from a memory page associated with data stored on the one or more tiers of memory, sampling the portion of raw data to select a sample data, analyzing the sample data to determine a sample category, and classifying the portion of raw data based at least in part by considering the sample category.

Abstract: Social network postings, including text, images or other media, may provide valuable information regarding a user of the social network with which the postings may be associated. With the authorization of the user, and upon authentication by the social network, an online marketplace may access the social network postings and extract data therefrom, and market one or more recommended items to the user based on the extracted data, which may include color pallets or texture pallets derived from photographs included in the postings.

Abstract: An information processing apparatus acquires an image obtained by photographing a part of the subject, specifies a relative position which part of the subject is photographed to obtain the image, determines a specific area included in the obtained image based on the specified relative position, and calculates an edge amount of the specific area. Then, in a case where the calculated edge amount in the specific area is larger than a first threshold, a character recognition processing is performed on the specific area included in the acquired image.

Abstract: A system is disclosed that obtains an image of an apparel item and text describing the apparel item. The system also determines descriptive metadata of the apparel item by extracting information from the text. Further, the system further determines categorization metadata of the apparel item by analyzing the image. Additionally, the system stores the image, the descriptive metadata, and the categorization metadata in a virtual closet. Moreover, the system provides a recommendation of an outfit from a stylist to a user based on the virtual closet.

Abstract: An image processing apparatus includes a computer system comprising an image acquiring section that acquires an image of body tissue. An extracting section of the computer system extracts an outline of the body tissue from the image, and a converting section of the computer system converts a coordinate sequence of the outline into a value sequence. An estimating section of the computer system estimates an abnormal shape of the body tissue by performing neural network processing on the value sequence. In addition, present invention embodiments include a computer program product used by the image processing apparatus and an image processing method performed by the image processing apparatus.

Abstract: The disclosure is related to a method and a system for estimating a stock level of a shelf. An image processing technique is incorporated to conduct stock estimation that is used as a reference for replenishment decisions. In the method, a camera is used to capture a shelf image. The shelf image is divided into multiple sub-region images. Multiple binarized images with respect to the multiple sub-region images are obtained by applying a binarization method. A ratio of the binarization values with respect to the pixels within every sub-region is calculated. The method also includes comparing the ratio of the binarization values in at least two sub-regions with a reference ratio in the at least two sub-regions within a shelf reference image. The comparison is used to estimate the stock level, and also to determine whether the shelf is to be replenished.

Abstract: A system and method for inline object detection using hue saturation value. One method includes determining, with an electronic processor running a single object classifier, a hue saturation value range. The method includes receiving a digital image including an object. The method includes detecting, without reloading the single object classifier, a macroblock from the digital image, the macroblock associated with the object. The method includes determining a target region within the macroblock. The method includes determining a quantity of pixels in the target region having a hue saturation value within the hue saturation value range. The method includes, when the quantity of pixels exceeds a threshold, completing object classification of the macroblock.

Abstract: A system and method are provided for brain tissue classification, which involves applying an automated tissue classification technique to an image of a brain based on a prior probability map, thereby obtaining a tissue classification map of the brain. A user is enabled to, using a user interaction subsystem, provide user feedback which is indicative of a) an area of misclassification in the tissue classification map and b) a correction of the misclassification. The prior probability map is then adjusted based on the user feedback to obtain an adjusted prior probability map, and the automated tissue classification technique is re-applied to the image based on the adjusted prior probability map. An advantage over a direct correction of the tissue classification map may be that the user does not need to indicate the area of misclassification or the correction of the misclassification with a highest degree of accuracy. Rather, it may suffice to provide an approximate indication thereof.

Abstract: A three dimensional object recognition device has in-vehicle cameras capturing front images of a vehicle, a recognition section and a light shaft judgment section. The recognition section detects a three dimensional object in the front images based on a parallax generated between the front images. The light shaft judgment section detects whether a light shaft phenomenon has happened in a three dimensional object area in at least one of the front images, and judges that the detected three dimensional object is an incorrect three dimensional object caused due to the light shaft phenomenon when detecting the light shaft phenomenon at every time when the recognition section detects the three dimensional object.

Abstract: One embodiment provides an image processing circuitry. The image processing circuitry includes a feature extraction circuitry and an optimization circuitry. The feature extraction circuitry is to determine a feature descriptor based, at least in part, on a feature point location and a corresponding scale. The optimization circuitry is to optimize an operation of the feature extraction circuitry. Each optimization is to at least one of accelerate the operation of the feature extraction circuitry, reduce a power consumption of the feature extraction circuitry and/or reduce a system memory bandwidth used by the feature extraction circuitry.

Abstract: Systems, methods, and computer program products for proactively improving image analysis and/or processing in real time are presented, in accordance with various embodiments. In one embodiment, a computer-implemented method includes: receiving an analysis and/or processing request from a mobile device; and either generating or retrieving analysis and/or processing information based on a result of either or both of: determining whether the analysis and/or processing request includes image data suitable for generating an analysis and/or processing result; and determining whether the analysis and/or processing request includes metadata suitable for generating an estimated analysis and/or processing result.

Abstract: A digital medium environment includes an image processing application that performs object segmentation on an input image. An improved object segmentation method implemented by the image processing application comprises receiving an input image that includes an object region to be segmented by a segmentation process, processing the input image to provide a first segmentation that defines the object region, and processing the first segmentation to provide a second segmentation that provides pixel-wise label assignments for the object region. In some implementations, the image processing application performs improved sky segmentation on an input image containing a depiction of a sky.

Abstract: Systems and methods for tracking moving objects in accordance with embodiments of the invention are disclosed. In one embodiment of the invention, an object tracking system comprises a processor, a communications interface, and a memory configured to store an object tracking application. The object tracking application configures the processor to receive a sequence of images; estimate and subtract background pixel values from pixels in a sequence of images; compute sets of summed intensity values for different per frame pixel offsets from a sequence of images; identify summed intensity values from a set of summed intensity values exceeding a threshold; cluster identified summed intensity values exceeding the threshold corresponding to single moving objects; and identify a location of at least one moving object in an image based on at least one summed intensity value cluster.

Abstract: Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for enforcing the category diversity or sub-category diversity of POIs that are identified in response to a local search. According to one implementation, a method includes receiving a request to identify points of interest (POIs), obtaining data identifying (i) candidate points of interest (POIs) that satisfy the request, (ii) a respective category associated with each candidate POI, and (iii) a non-scaled score associated with each candidate POI, and ranking, for each of one or more of the categories, the candidate POIs associated with the category, based on the respective non-scaled scores.

Abstract: Techniques are provided that include obtaining a Document Object Model of an HTML document, such as a web page of a publisher. Elements of the Document object model may be identified that are associated with native advertisement placement candidate containers. Based at least in part on analysis associated with the Document Object Model, and utilizing at least some of the identified elements, one or more native advertisement placement candidate containers may be determined. Some techniques may utilize, in the analysis, identification of one or more secondary source HTML documents associated with a primary HTML document, and may identify, within the secondary source HTML document(s), content associated with native advertisement placement candidate containers of the primary HTML document(s).

Abstract: A computer-implemented method includes receiving surveying data from a proximity barrier survey device. The proximity barrier survey device is mounted to a package. The package is located within a physical environment. The method includes generating a map based on the surveying data. The method includes tracking movement within the physical environment based on the map and the surveying data.

Abstract: A complete personal nutrition, health, wellness and fitness monitor is disclosed that captures, monitors, and tracks many relevant health and wellness factors. Image of all consumed items is captured with a three dimensional reference object, matched to reference images in a database in terms of shape, color, size and texture, identified, and the volume of the respective items is determined. In addition, molecular optical fingerprinting of consumable products provides information about actual food composition, including vitamins and contaminants that affect food safety. Using energy and nutrition information of items in another database, full nutritional content and energy is evaluated for all of the items in the image.

Abstract: A method and system for conducting image search is provided that includes searching a first database to locate a set of pertinent images, iteratively performing the following operations for each image in the set of pertinent images (a) extracting the histogram of red green and blue colors (RGB colors) from a given image (b) distilling the extracted RGB colors down to create a reduced color palette for the given image (c) segmenting the extracted RGB colors into a set of segments representing distinct parts of the color spectrum; (d) selecting a subset of segments to assemble a color palette for the given image; (e) updating the assembled color palette and the customer behavior score for the given image in the first database; and generating a ranked display of suggested color palettes for the search query by aggregating the customer behavior scores for the search query across all images.

Abstract: A method for mapping a sensor pixel array to an illumination pixel array according to a surface forms a group mapping by assigning each pixel in a on the sensor array to a group that has a group width defined by p adjacent pixels on the illumination pixel array by projecting and recording a first and a second multiple group index image with a first and second pattern of lines. Lines appearing in both first and second pattern are spaced by a first distance that is a first multiple of group width p, and lines that appear only in either pattern are evenly spaced by a second distance that exceeds the first distance. A subset of p multiline images are projected, each projecting a line within each group. Lines in one of the multiline images are correlated with lines from the group index images to generate the group mapping.

Abstract: A method of modifying a scene model for a sequence of input images includes receiving, selecting, determining, and modifying. Scene modes associated with the sequence are received. For each image pixel in an input image of the sequence, two scene modes are selected from the received scene modes based on distances between the image pixel color value and the scene modes color values at a corresponding pixel location. It is determined if the image pixel color value matches the color value of any of the selected scene modes. It is determined if the image pixel color value is a mixture of the color values of the selected scene modes. The color value of a closest selected scene mode is modified or is maintained depending on whether there is one matched mode and the image pixel color value is not a mixture of the color values of the selected scene modes.

Abstract: An activity recognition system is disclosed. A plurality of temporal features is generated from a digital representation of an observed activity using a feature detection algorithm. An observed activity graph comprising one or more clusters of temporal features generated from the digital representation is established, wherein each one of the one or more clusters of temporal features defines a node of the observed activity graph. At least one contextually relevant scoring technique is selected from similarity scoring techniques for known activity graphs, the at least one contextually relevant scoring technique being associated with activity ingestion metadata that satisfies device context criteria defined based on device contextual attributes of the digital representation, and a similarity activity score is calculated for the observed activity graph as a function of the at least one contextually relevant scoring technique, the similarity activity score being relative to at least one known activity graph.

Abstract: A method for evaluating quality of tone-mapping image based on exposure analysis is provided, which explores the exposure properties on each area of the high dynamic range image utilizing the pre-exposure method and divides the high dynamic range image into three parts of an easy overexposed area, an easy underexposed area and an easy natural-exposed area, wherein different quality characteristics are extracted in different areas, which is capable of ensuring that the follow-up quality characteristic extraction is more targeted. The present invention takes the difference of distortion between the tone-mapping image and the conventional image into account, and extracts image characteristics such as the abnormal exposure rate, the underexposed residual energy, the overexposed residual energy and the exposure color index, so as to accurately reflect the quality degradation of the tone-mapping image.

Abstract: An apparatus and method to adjust item recommendations are disclosed herein. A first image attribute of a query image is compared to a second image attribute of each of a plurality of inventory images of a plurality of inventory items to identify the inventory items similar to the query image. Item recommendations comprising the identified inventory items in a first listing order are provided for display at a remote device. A second listing order of the identified inventory items is determined based on a user preference for a particular one of the identified inventory items. At least the second listing order is provided to the remote device for re-display of the item recommendations in accordance with the second listing order.

Abstract: A method is provided for detecting static graphical elements in a sequence of video frames that compares a selected frame in the sequence to each of a plurality of previous frames in the sequence to identify a graphical element that can be a logo. For each pair of frames compared, an absolute difference frame is determined by acquiring an absolute difference value between pixel values for corresponding pixels over at least a portion of the frames in the frame pair. A metric associated with each absolute difference frame is generated, which reflects a degree of dissimilarity. At least some of the absolute difference frames weighted in accordance with the metric associated therewith are summed to generate an accumulation difference frame such that pairs of frames that are more dissimilar have a greater weight. A static graphical element is then identified over a region of the accumulation difference frame in which pixel values satisfy specified criteria.