Abstract: An image processing method and device, an equipment, and a computer-readable storage medium are provided. The method includes the following. An image is obtained. Object detection is performed on the image with an object detection network to obtain an object detection result. The object detection network is a network obtained by performing object detection on sample images in a sample image set based on at least one cluster center value with an initial object detection network and training iteratively the initial object detection network based on results of the object detection. The at least one cluster center value is obtained by clustering labeled regions in a labeled region set corresponding to the sample image set based on preset attribute information. The image is processed based on the object detection result.

Abstract: 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.

Abstract: A spectral imaging device includes an imager, a scanning stage to establish relative motion between the imager and a sample in a scanning direction and an optical system controlling a light characteristic of a light beam constituting an image of the sample to the imager. The optical system includes a light varying element to receive the light beam and provide an output light beam with spatially varying light characteristic over a cross-section thereof. A set of redirecting optical elements direct light rays from the sample to form the light beam, and to focus the output light beam onto the imager. A controller controls the scanning stage and the imager to capture a plurality of image frames with an overlap including a defined shift that is greater than 1 pixel along the scanning direction between consecutive image frames. A computing device consolidates image data to provide an image of the sample.

Abstract: A method of detecting a stain of a display panel is disclosed that includes capturing a first color image of the display panel, capturing a second color image of the display panel, generating a merged image by merging a color coordinate map of the first color image and a color coordinate map of the second color image, generating a background image of the merged image by a morphology filtering, generating a flattened image by operating the merged image and the background image and detecting an area of the flattened image in which a color coordinate value exceeds a threshold value as a stain area.

Abstract: Devices, methods, and processor-readable media for eye tracking using statistically derived linear functions are described. Two simple linear functions may be identified, one for the horizontal and one for the vertical, with two parameters each, that fully correct for horizontal and vertical estimation bias error in a cross-ratio gaze estimation method. An implicit calibration step may be performed that finds the parameters of said linear functions statistically based on undirected user point of gaze measurements that indicate the extremities of the error between gaze vectors and glint vectors. Various hardware devices are described that may implement the novel eye tracking techniques.

Abstract: A data processing system implements a receiving a textual input comprising a query for a first image. The data processing system also implements analyzing the textual input to determine a predicted color palette associated with a subject matter of the query; and procedurally generating the first image using the predicted color palette. Another implementation of the data processing system implements providing the textual input to a first machine learning model to obtain the first image, the first machine learning model being trained using a dataset comprising abstract imagery and analyzing the textual input using the first machine learning model to obtain the first image in response to receiving the textual input.

Abstract: Techniques are described to generate a 3D scene by mapping a point cloud with a 2D image, and colorize portions of the 3D scene synthetically. An input is received to select, from the 3D scene, a portion to be colorized synthetically. The colorizing includes generating a reflectance image based on an intensity image of the point cloud. The colorizing further includes generating an occlusion mask that identifies the selected portion in the reflectance image. The colorizing further includes estimating, using a trained machine learning model, a color for each of the one or more points in the selected portion based on the reflectance image, the occlusion mask, and the 2D image. The 3D scene is updated by using the estimated colors from the trained machine learning model to colorize the selected portion.

Abstract: The disclosed technique detects a foreground region accurately even with a neural network. A first region detection unit detects, by a DNN, a predetermined foreground region in an inputted image. A weak region identification unit detects, by a DNN, a weak region having a possibility that misdetection or non-detection occurs during the region detection on the inputted image using a DNN that has the same detection target as that of the DNN of the first region detection unit. A second region detection unit detects, by a method other than a neural network, a foreground region that is a detection target same as that of the first region detection unit in the weak region detected by the weak region identification unit within the inputted image read from an input device. An integration unit integrates detection results by the first region detection unit and by the second region detection unit.

Abstract: Method and apparatus for recommending at least one of a makeup palette or a hair coloration scheme are provided. The method comprises extracting a color set of at least one region of a digital image associated with a user (201); generating color recommendation information for at least one of a makeup palette or a hair coloration scheme for at least two other regions of the digital image associated with the user based on one or more inputs indicative of the extracted color set (202); and generating one or more instances of a virtual representation of a makeup palette recommendation or a hair coloration scheme recommendation based on the color recommendation information (203).

Abstract: An apparatus for determining an index of insect biodiversity comprises a plurality of optical insect sensor devices configured to be individually positioned within a geographic area and a data processing system. Each sensor device is configured to monitor insect activity within a detection volume by detecting light from the detection volume, and to output detector data indicative of one or more optically detected attributes associated with respective detected insect detection events. Each insect detection event is indicative of one or more insects being present in the detection volume. The data processing system is configured to receive detector data from respective ones of the plurality of optical insect sensor devices. The detector data is indicative of one or more optically detected attributes associated with respective detected insect detection events. The system computes, from the received detector data, an index of insect biodiversity indicative of insect biodiversity within the geographic area.

Abstract: Discussed is a customized cosmetics provision system including a diagnosis device for diagnosing a skin condition, recommending customized cosmetics based on a diagnosis result of the skin condition, and transmitting information on the recommended customized cosmetics to a manufacturing apparatus, and the manufacturing apparatus that receives information on the customized cosmetics from the diagnosis device and provides the customized cosmetics based on the received information on the customized cosmetics. The diagnosis device can output a message to re-diagnose the skin condition when an environmental condition for diagnosing the skin condition do not meet a predetermined condition.

Abstract: Disclosed are a system and process of providing social networking purchasing processes. A method includes receiving, from a posting entity and at the social networking entity, a posting or a submission of a text, an image or a video. When the submission is associated with a product within a product catalog of the posting entity, the social networking entity presents in a newsfeed of users or otherwise on the social networking entity, the text, image or video with an associated option to buy. The option to buy might be presented through a messenger application or as the user browses the posting. When there is a correlation between the posting and the product catalog, and when the user makes a purchase of the product, the user is not transitioned away from the social networking entity. Initiating a payment process occurs within the social networking entity.

Abstract: The disclosure relates to systems, methods and programs for maneuvering unmanned vehicles. More specifically, the disclosure relates to systems, methods and programs for controlling maneuverability of unmanned vehicles (ground, aerial and marine) by coupling vehicle controls with point of regard (PoR) in a 2D plane, translated to a continuously updating flight vector in a 3D space, based on 12 DOF head pose and/or hand gesture of a user.

Abstract: Disclosed is an interactive product shopping method including acquiring a real-time broadcast image related to the sale of a target product, acquiring product information of the target product, acquiring information on a sales method of the target product, acquiring input information and identification information of a candidate purchaser who requests to purchase the target product, determining a prospective purchaser of the target product on the basis of a quantity information of the target product and the input information of the candidate purchaser, acquiring shopping cart information of the prospective purchaser on the basis of the identification information of the prospective purchaser and assigning the product information of the target product to a shopping cart of the prospective purchaser, and acquiring payment intention information of the prospective purchaser and updating the sales status of the target product on the basis of the payment intention information.

Abstract: Techniques for reducing the latency of annotating a replay video segment may include receiving a video segment with content involving multiple individuals. An annotation task is performed concurrently with a tracking task. The annotation task receives annotation data to indicate which of the individuals is an individual of interest in a subset of frames of the video segment, and the tracking task tracks the individuals in the video segment by generating bounding objects corresponding to the individuals. The annotation data can be associated with the bounding objects to detect a bounding object for the individual of interest, and a visualization emphasis object is generated based on the detected bounding object in a replay video segment to identify the individual of interest.

Abstract: A character recognition method includes the stages as detailed in the following paragraph. An image is received, wherein the image is one in a plurality of consecutive images. A target object in the image is detected. Object information of the target object is defined according to the area ratio of the target object occupied in the image. Whether the target object in the image is the same as the target object in the previous image is determined according to the object information. Character recognition on the target object is performed to obtain a recognition result. The weighting score of the recognition result is calculated according to the object information and the recognition result. The weighting score of the recognition result of the target object in the consecutive images is accumulated until the weighting score is higher than a preset value, and the recognition result is output.

Abstract: One variation of a method for stock keeping in a store includes: accessing an image captured by a fixed camera within the store; retrieving a field of view of the fixed camera; estimating a segment of an inventory structure in the store depicted in the image based on a projection of the field of view onto a planogram of the store; identifying a set of slots within the inventory structure segment; retrieving a product model representing a set of visual characteristics of a product type assigned to a slot, in the set of slots, by the planogram; extracting a constellation of features from the image; if the constellation of features approximates the set of visual characteristics in the product model, detecting presence of a product unit of the product type occupying the inventory structure segment; and representing presence of the product unit, occupying the inventory structure segment, in a realogram.

Abstract: In a data theft prevention method, a data processing apparatus obtains target data, inputs the target data into a recognition model for recognition processing to obtain a first label, where the first label is a label to which the target data belongs. The recognition processing includes: performing, by the recognition model, feature extraction on the target data to obtain a first feature vector; and obtaining, by the recognition model, the first label through prediction based on the first feature vector; and outputs a target label when determining that a second feature vector exists. The second feature vector is a feature vector whose distance from the first feature vector is less than a first threshold. A second label obtained through prediction by using the recognition model is different from the first label and the target label is a label obtained by performing perturbation processing on the first label.

Abstract: An apparatus for evaluating color of an object includes an acquisition unit configured to acquire image data which is acquired by imaging the object and a color chart group including color charts having different reflectance within a one angle of view, an estimation unit configured to estimate a value of reference white under an exposure condition corresponding to the imaging based on the image data, and a calculation unit configured to calculate a color value for evaluating the color of the object based on the image data and the estimated value of the reference white.

Abstract: An image capture apparatus obtains image data captured using an image sensor and then generates, as metadata, information relating to a state of the image capture apparatus at the time of capturing the image data as well as to the image data. The image capture apparatus comprises a first output circuit for outputting the image data in a first format to an external apparatus and a second output circuit for outputting a part of the metadata in a second format to the external apparatus. The second output circuit is from the first output circuit.

Abstract: Systems and methods are provided for encoding a frame of 3D media content. The systems and methods may be configured to access a first frame of 3D media content and generate a data structure for the first frame based on color attributes information of the first frame, wherein each element of the data structure encodes a single color. The systems and methods may be configured to train a machine learning model based on the first frame of 3D media content, wherein the machine learning model is trained to receive as input a coordinate of a voxel of the first frame, and to output an identifier of a particular element in the generated data structure. The systems and methods may be configured to generate encoded data for the first frame based at least in part on weights of the trained machine learning model and the generated data structure.

Abstract: Disclosed herein are systems, methods, and computer-readable storage devices for providing payment data from one device to another via a payment protocol. A device establishes, via a wireless link, a communication between the device and a user device The device transmits, via a protocol a signal to the user device to initiate a payment process on the user device to make the purchase. The signal is configured to cause the user device, automatically and without any other user interaction, to next receive a confirmation from the user to finalize the purchase. The device receives, via the wireless link, from the user device and at the wireless communication device, user payment data to make the purchase. The user payment data is obtained based on the confirmation and other parameters. The device transmits, to the user device and via the wireless link, a payment confirmation upon completion of the payment process.

Abstract: Images of an insect are subjected to at least a first convolutional neural network to develop feature maps based on anatomical pixels at corresponding image locations in the respective feature maps. The anatomical pixels correspond to a body part of the insect. A computer calculates an outer product of the first feature map and the second feature map to form an integrated feature map. Extracting fully connected layers from respective sets of integrated feature maps and applying the fully connected layers to a classification network for identifying the genus and the species of the insect.

Abstract: A search support system includes at least one camera, and a server capable of communicating with the camera. The camera transmits, to the server, a captured image of the object and image color information of the captured image. The server determines whether a search condition of the captured image set by a user includes a color condition of at least one of white, black, and gray. When the color condition is included in the search condition, the server excludes a captured image associated with the image color information indicating the black-and-white image among all captured images stored in the database from a search target, and outputs a search result including a captured image satisfying the search condition. When the color condition is not included in the search condition, the server outputs a search result including a captured image satisfying the search condition among all captured images stored in the database.

Abstract: Methods and systems for analyzing a field. The methods and systems acquire a thermal image indicative of thermal energy emitted by the soil and/or plants in the field and process the thermal image to assess variations in certain characteristics of the soil and/or plants.

Abstract: Systems, devices, and methods related to a Deep Learning Accelerator and memory are described. For example, a data storage device may be configured to execute instructions with matrix operands and configured with: an interface to receive a video stream; and random access memory to buffer a portion of the video stream as an input to an Artificial Neural Network and to store instructions executable by the Deep Learning Accelerator and matrices of the Artificial Neural Network. The Deep Learning Accelerator can execute the instructions to generate an output of the Artificial Neural Network, including analytics of the buffer portion. A video encoder in the data storage device may use the analytics to compress the portion of the video stream for storing in the device.

Abstract: A system and method for providing payments is disclosed. A method can include establishing a wireless link between a mobile device and a merchant device. After an instruction is displayed on the mobile device, a combination of a first type of input and a second type of input are received on the mobile device, first type of input including at least one button press of a physical button and the second type of input received from a user to confirm a payment for the purchase. Based on the inputs, payment data is retrieved from a memory of the mobile device. The method includes receiving the payment data at the merchant device to make the purchase, wherein receiving the payment data, via the wireless link, at the merchant device to make the purchase is performed according to a protocol for communicating the payment data to the merchant device.

Abstract: A system and method are disclosed that allow for early detection and management of wildfires. According to various aspects and embodiments of the invention the system and method rapidly and economically detect and help manage and contain, control, suppress the progress of wildfires.

Abstract: This disclosure relates to an image segmentation method and apparatus and image three-dimensional reconstruction method and apparatus, and to the field of computer technology. The segmentation method includes: dividing pixels in an image to be segmented among different pixel sets, according to color gamut ranges to which pixel values of the pixels belong; determining matching between pixels in each pixel set according to pixel values; and performing image segmentation on the image to be segmented according to the matching.

Abstract: A method and an apparatus for detecting key points of a human body, and a storage medium are provided. The method includes: extracting, in response to detecting that an image comprises a body, two-dimensional (2D) coordinate data for identifying body keypoint positions in the image to obtain 2D pose data; and performing body keypoint feature fusion on the 2D pose data and depth data corresponding to the body keypoint positions to obtain three-dimensional (3D) pose data for identifying the body keypoint positions.

Abstract: An automated inventory monitoring system includes an image capture module able to create an image of an aisle of a retail store. Images of products labels are identified in the image and classified as shelf labels or peg labels. For shelf labels, an area of the shelf is defined and associated with the shelf label. Images of products are identified in the image and products on the shelf within an area associated with a shelf label are associated with the shelf label. Products located below a peg label are associated with the peg label. Based on the association between labels and products, out-of-stock products, plugs and spread may be detected and reported to the staff of the retail store.

Abstract: A device may receive receipt information associated with items in an order. The receipt information may include item identifiers and price descriptors associated with the items in the order. The device may receive product information associated with the items in the order. The product information may include product identifiers based on images of the items in the order and object descriptors. The device may map a first item identifier for a first item in the order and a first product identifier for determining that the first item corresponds to a first price descriptor and a first object descriptor. The device may map a second item identifier for a second item in the order and a second product identifier for determining that the second item corresponds to a second price descriptor and a second object descriptor. The device may assign the first and second items to respective first and second sub-receipts.

Abstract: Systems, devices, and methods related to a deep learning accelerator and memory are described. For example, a removable media (e.g., a memory card, or a USB drive) may be configured to execute instructions with matrix operands and configured with: an interface to receive a video stream; and random access memory to buffer a portion of the video stream as an input to an artificial neural network and to store instructions executable by the deep learning accelerator and matrices of the artificial neural network. Such a removable media can be used to replace an existing removable media used in a surveillance camera to record video or images. The deep learning accelerator can execute the instructions to generate analytics of the buffer portion using the artificial neural network, enabling the surveillance camera that is upgraded via the use of the removable media to provide intelligent services based on the analytics.

Abstract: A vehicular vision system includes a camera disposed at a vehicle and capturing image data. The system, via processing at an electronic control unit (ECU) of image data captured by the camera, detects an object at a first orientation relative to camera. A first local binary pattern represents in binary form a first portion of the image data that includes the detected object. The system detects the object at a second orientation relative to camera. A second local binary pattern represents in binary form a second portion of the image data that includes the detected object. The second orientation is different from the first orientation and the second local binary pattern is different than the first local binary pattern. The system groups the first and second local binary patterns into a common histogram bin. The system classifies the detected object based at least in part on the common histogram bin.

Abstract: A method and an apparatus for detecting key points of a human body, and a storage medium are provided. The method includes: extracting, in response to detecting that an image comprises a body, two-dimensional (2D) coordinate data for identifying body keypoint positions in the image to obtain 2D pose data; and performing body keypoint feature fusion on the 2D pose data and depth data corresponding to the body keypoint positions to obtain three-dimensional (3D) pose data for identifying the body keypoint positions.

Abstract: A system and method for providing payments is disclosed. A method can include establishing a wireless link between a mobile device and a merchant device. After an instruction is displayed on the mobile device, a combination of a first type of input and a second type of input are received on the mobile device, first type of input including at least one button press of a physical button and the second type of input received from a user to confirm a payment for the purchase. Based on the inputs, payment data is retrieved from a memory of the mobile device. The method includes receiving the payment data at the merchant device to make the purchase, wherein receiving the payment data, via the wireless link, at the merchant device to make the purchase is performed according to a protocol for communicating the payment data to the merchant device.

Abstract: Technology is described for generating a modified logo using popular colors of a logo file to be used in electronic promotional pages. In addition, options selected by a user, the color of the logo itself and attributes of the electronic promotional pages can be used to filter electronic promotional pages provided. One method may be the operation of detecting a hue, saturation and luminance (HSL) in pixels of a logo file. Another operation may be clustering of the pixels from the logo file into groups based on hue, saturation and luminance (HSL). The colors in the color clusters may be weighted based on area of the logo file occupied by a color. Colors that exceed a defined weighting percentage may be selected as a popular color. Any repeated popular colors may be removed. Then a modified logo may be generated using the popular color selected.

Abstract: In implementations of three-dimensional vector-based brushes, a computing device implements a brush system to receive input data describing a first stamp and a second stamp of consecutive stamps of a vector-based digital brush. The first stamp is segmented into a first convex geometry and a second convex geometry, and the second stamp is segmented into a third convex geometry and a fourth convex geometry. The brush system computes a first convex hull of the first convex geometry and the third convex geometry and computes a second convex hull of the second convex geometry and the fourth convex geometry. An order for passing the first convex hull and the second convex hull to a planarizer is determined and the brush system generates a vector-based stroke of digital paint for display in a user interface based on the order.

Abstract: A method to improve the efficacy of optical character recognition (OCR) includes scanning an electronically stored representation of a whole or partial document, identifying an image having text in the electronically stored representation of a whole or partial document, identifying the text within the image, and generating a plurality of bounding boxes around the identified text using blob detection. The method also includes grouping together certain text bounding boxes of the plurality of text bounding boxes that are vertically aligned with each other to generate a plurality of aligned text bounding boxes and performing OCR on the aligned text bounding boxes to generate a plurality of OCR groups of text. In addition, the method includes generating a resultant representation of a whole or partial document electronically using the plurality of OCR groups of text and saving the resultant representation of a whole or partial document electronically.

Abstract: Disclosed are a system and process of providing social networking purchasing processes. A method includes receiving, from a posting entity and at the social networking entity, a posting or a submission of a text, an image or a video. When the submission is associated with a product within a product catalog of the posting entity, the social networking entity presents in a newsfeed of users or otherwise on the social networking entity, the text, image or video with an associated option to buy. The option to buy might be presented through a messenger application or as the user browses the posting. When there is a correlation between the posting and the product catalog, and when the user makes a purchase of the product, the user is not transitioned away from the social networking entity. Initiating a payment process occurs within the social networking entity.

Abstract: The present invention relates to a method for classifying a fundus image and a device using same. Specifically, according to the method of the present invention, a computing device acquires a fundus image of a subject, generates classification information of the fundus image, generates an interpretation text on the basis of the classification information, and provides the interpretation text to an external entity.

Abstract: A display control apparatus includes a change unit configured to change amount of application of a predetermined effect which is to be applied to a selected subject from a plurality of subjects, and a control unit configured to perform control to display a first item which makes the selected subject distinguishable from other subjects, wherein, in response to the change unit starting changing amount of application of the predetermined effect, the control unit performs control not to display the first item.

Abstract: An illumination apparatus irradiates first reference light having a first intensity distribution and second reference light having a second intensity distribution that has a complementary relation with the first intensity distribution. A photodetector measures reflected light from an object. A processing device executes a first correlation calculation using a first detection intensity based on the output of the photodetector in a state in which the first reference light is irradiated and the second intensity distribution, and executes a second correlation calculation using a second detection intensity based on the output of the photodetector in a state in which the second reference light is irradiated and the second intensity distribution.

Abstract: There are provided systems and methods for dynamically rendered interface elements during online chat sessions. A user may engage in online communications with another user, such as a communication session between a customer and a customer representative or agent of a merchant. During this communication session, the agent may navigate to particular data on the agent's device, which may be desirable to provide to the customer. For example, the agent may view a particular item of interest to the customer. An application programming interface of the merchant may detect one or more actions or calls associated with this data and may dynamically provide an interface element to transmit data displayed on the agent's device to the customer's device during the communication session. The customer may view a dynamically rendered interface element that allows for processing data during the communication session with the agent.

Abstract: Method and apparatus are disclosed for collective electronic commerce and distribution of items, such as fungible items/goods and non-fungible items/goods. A system manages a universal ecommerce shopping cart to facilitate creating an itinerary for a group across multiple vendor websites/applications. The systems manages inventory to interface with multiple vendors, manages groups, and/or manages the assignment of fungible items (e.g., products) and non-fungible items (e.g., tickets, access to hotel rooms, etc.) that were selected for being purchased at different times.

Abstract: Disclosed is a method including receiving at a user device a scan of a code, initiating on the device, based on the code, a communication with a server, receiving, based on the communication and from the server, a download of a user interface, the download being independent of a full application stored on the user device, transmitting, via the software module, an identification, received from the user, of a function to perform in connection with the code, transmitting location data associated with the user device to the server and transmitting authorized data to the server based on data retrieved from either the user device or a network entity. The user can scan product codes or rental codes in the store or on a device like a scooter and receive the interactive interface for performing a function. The process eliminates the need for point of sale equipment and eliminates checkout lines.

Abstract: There are provided an image processing device, an image processing method, a program, and a portable terminal device which more accurately extract a user's region of interest from an image with fewer operations. The object is solved by an image processing device which divides an acquired image into a plurality of regions, selects at least one region from among the plurality of divided regions, analyzes a relationship between the selected region and a region other than the selected region, and outputs a grouping candidate for a region belonging to the same group as the selected region from the analyzed relationship.

Abstract: A system for identifying products in a retail store is provided. The system may include a processor configured to: receive an image depicting a store shelf and a product displayed thereon; select a product model subset based on a characteristic of the store shelf determined based on analysis of the image; determine whether the product model subset is applicable to the product; and when the product model subset is applicable to the product, analyze a representation of the product depicted in the image using the product model subset, and identify the product based on the analysis of the representation of the product.

Abstract: The technology described herein visibly depicts hidden message traits to help users determine whether an email is genuine or deceptive. The hidden message traits are revealed by identifying and changing attributes that keep the hidden traits from being displayed in a rendered message. Spam messages, phishing messages, and messages that include or link to malicious programs (e.g., malware, ransomware) are examples of unwanted messages that can harm a recipient. These messages often rely on deception to get past email filtering systems and to trick a user into acting on content in a message. The deception often involves including hidden traits in a message that fool an automated filtering system. The technology described herein shows the visible traits to a user by including them in the rendered version of the message.

Abstract: Disclosed is a web 3.0 orchestration system that creates a token economy through which customers are able to establish relationships and purchase physical and digital goods and services from any merchant at any place without having to access the physical or digital storefront of the merchant. The orchestration system may receive a triggering event from a user device, may compile user data from user profiles that are associated with the user device, and may select a distributed set of smart contracts defined by different merchants based on the triggering event. The orchestration system may execute the distributed set of smart contracts using the user data as inputs, may select offerings from a set of merchants based on the output generated from executing the smart contracts, and may present the offerings to the user device as part of facilitating a transaction between the user device and any of the merchants.