Abstract: Improving accuracy in understanding and/or resolving references to visual elements in a visual context associated with a computerized conversational system is described. Techniques described herein leverage gaze input with gestures and/or speech input to improve spoken language understanding in computerized conversational systems. Leveraging gaze input and speech input improves spoken language understanding in conversational systems by improving the accuracy by which the system can resolve references—or interpret a user's intent—with respect to visual elements in a visual context. In at least one example, the techniques herein describe tracking gaze to generate gaze input, recognizing speech input, and extracting gaze features and lexical features from the user input. Based at least in part on the gaze features and lexical features, user utterances directed to visual elements in a visual context can be resolved.

Abstract: An image recognition method of a terminal is provided. An image sensor is controlled to capture an image stream in response to receiving an image recognition trigger. A first image is extracted from the image stream. A determination is made as to whether an image type of the first image is one of an object image and a code image. One of a plurality of recognition processes is performed on a second image in the image stream according to the determined image type.

Abstract: A digital image can be watermarked by a computer-implemented method. An image analysis process can be performed on the digital image to identify an area of the digital image containing at least one predetermined object. Watermark information representative of content of the identified area can be generated, and a digital watermark containing the generated watermark information can be subsequently watermarked/overlaid onto the digital image.

Abstract: In some examples, natural language processing based sign language generation may include ascertaining a speech video that is selected by a user, and determining, based on application of natural language processing to contents of the speech video, a plurality of sentences included in the speech video. For each sentence of the plurality of sentences identified in the speech video, a sign language sentence type, a sign language sentence structure, and a sentiment may be determined. For each sign language sentence structure and based on a corresponding sentiment, a sign video may be determined. Based on the sign video determined for each sentence of the plurality of sentences identified in the speech video, a combined sign video may be generated.

Abstract: The disclosure relates to detecting digital watermarking from retail items such as from packaged items, containers, bottles, cans or boxes. One claim recites a method utilized at a retail checkout location comprising: receiving imagery representing a retail item from a digital camera, the retail item including digital watermarking encoded thereon, the retail item moving relative to the digital camera; determining a region in the imagery corresponding to at least one faster moving object relative to background imagery, said determining yielding a determined region; arranging digital watermark detection blocks over the determined region; and analyzing data representing imagery from within the digital watermark detection blocks to detect the digital watermarking. Of course other claims and combinations are also provided.

Abstract: Disclosed herein is an apparatus for editing graphic objects. More specifically, disclosed herein is an apparatus for editing graphic objects which searches for objects by using a variety of attributes that each of the objects has and displays the found objects and a list of the found objects. The apparatus includes: a display unit configured to display a graphic area and a list area; an input unit configured to receive filter attributes associated with objects from a user; a search unit configured to search one or more objects for an object having object attributes matching the filter attributes; and a control unit configured to control the display unit so that the objects found by the search unit are displayed in the graphic area and a list of the objects found by the search unit is displayed in the list area.

Abstract: Included are an analysis unit that performs an analysis process for analyzing a specimen, a storage unit that stores a result of the analysis process, cumulative information on an occurrence of an event such as registration, change or deletion of analysis parameters such as analysis conditions in the analysis process, and cumulative information on an occurrence of an event such as registration, deletion or replacement of reagent information on a reagent used for analysis of the specimen, for each analysis item, and a display unit that displays the analysis result of the specimen and information on an occurrence status of the event. An operation of the analysis process of the specimen by the analysis unit is controlled, and the information on the event of the analysis parameter and the information on the event of the reagent information is read from the storage unit and displayed in chronological order.

Abstract: A method for securely accessing a hardware storage device connected to a computer system, the hardware storage device having a unique hardware identifier and the computer system including a processor, the method comprising: an agent software component receiving the identifier of the storage device to authenticate the storage device, wherein the agent executes in an unrestricted mode of operation of the processor such that the agent is a trusted software component; in response to the authentication, the agent accessing a secure data key for encrypting and decrypting data on the storage device, wherein the data key is accessible only to trusted agents executing in the unrestricted mode of the processor such that software executing in a user mode of the processor stores and retrieves data on the storage device only via the agent.

Abstract: A method in an access control system includes maintaining a database of access control event data generated by a plurality of access control devices installed in a building and iterating through the access control event data in order to generate a connectivity model for the building. Generating the connectivity model includes, for each iteration, identifying a user interaction with a first door at a first time and a user interaction with a second door at a second time, determining if a difference between the first time and the second time is less than a threshold period of time, and either creating the connection between the first door and the second door or updating a weight associated with the connection. The connectivity model can be used in the access control system to correlate access control events across locations in a building or building campus.

Abstract: A system includes circuitry configured to embed a digital watermark data in an original image, wherein the original image is divided into a plurality of blocks each block having been embedded with a pattern corresponding to each value of the digital watermark data, store, in a memory, the original image, the digital watermark data embedded in the original image, and an embedding position of the pattern of the digital watermark data in association, detect a pattern in a scanned image of a printed matter, decode the detected pattern to acquire digital watermark data included in the scanned image, align each block between the original image and the scanned image, based on the embedding position of the pattern associated with the original image and a detection position of the pattern detected from the scanned image, and obtain a difference between the original image and the scanned image aligned with each other.

Abstract: A method of authentication using object imaging, comprising: identifying at least one key geometric marker related to an object captured from a specific viewport and depicted in a reference image; receiving a current image of the object, the current image captured by an imaging device; extracting at least one current geometric marker related to the object from the current image; and comparing the at least one current geometric marker to the at least one key geometric marker to verify that the current image is depicting the object from the viewport.

Abstract: Provided is a printer including an ink tank having a mark at a side surface of the ink tank, a print head performing printing by using ink in the ink tank, a photoelectric conversion device provided outside the side surface having the mark among side surfaces of the ink tank, and detecting light from the ink tank, and a processing unit determining a position of an interface of the ink based on an output of the photoelectric conversion device and detecting an amount of ink in the ink tank based on a position of the mark and the position of the interface.

Abstract: There is provided a non-transitory computer readable storage medium tangibly embodying a machine-readable image having data encoded therein and embedded with a graphic, the machine-readable image adapted to be detected by a reader for decoding said encoded data, including: the graphic associated with an image descriptor calculated based on a chosen area of the graphic, the image descriptor being used in a reading process of said machine-readable image; a plurality of function patterns; and a plurality of dot modules having decoded values corresponding to at least said encoded data, the dot modules being positioned in one or more encoding regions of the machine-readable image relative to the function patterns and the chosen area of the graphic.

Abstract: A counterfeit detection system provides an artificial intelligence (AI) platform that implements a Generative Adversarial Network (GAN) to classify an image as one of a fake or genuine item and integrates a Classification Activation Module (CAM) to refine counterfeit detection. The GAN may include a generator that generates simulated counterfeit images for a discriminator. The discriminator may be trained to identify faked items by learning from the simulated counterfeit images and/or images of actual faked items. The discriminator may implement a deep neural network of convolutional layers that each analyze a region of an image and produce a weighted output that contributes to the classification based on the analyzed region. The CAM may identify the regions and weights relied upon by the discriminator, provide corresponding heatmaps to subject matter experts, receive annotations from the subject matter experts, and use the annotations as feedback to refine the classifier of the discriminator.

Abstract: To prevent deterioration in image quality in mixing a visible light signal and an infrared light signal to generate a color image signal. A signal processing device includes a mixture ratio creator and a mixer. The mixture ratio creator is configured to create a mixture ratio that is a ratio in mixing a low frequency visible light signal that is a low frequency component of a visible light signal and a low frequency invisible light signal that is a low frequency component of an invisible light signal together on a basis of the low frequency visible light signal and the low frequency invisible light signal. The mixer is configured to mix the low frequency visible light signal and the low frequency invisible light signal together on a basis of the mixture ratio to generate a mixed signal.

Abstract: A decoding device and a decoding method for decoding a target information from a bitstream of video data are provided, wherein the bitstream comprises a prediction parameter encoded in the bitstream. The decoding device is configured to calculate a randomized host signal based on the prediction parameter and a pseudo-random signal, and to apply a check function to the randomized host signal and thereby obtain the target information. An encoding device and an encoding method, and computer-readable storage media for encoding and for decoding are also disclosed.

Abstract: Disclosed herein is a posture control system containing a posture control unit that changes a direction of a user's body and a display unit mounted on a user's head. The posture control system includes: a posture data acquiring unit configured to acquire posture data indicating the direction of the user's body; a motion sensor data acquiring unit configured to acquire motion sensor data indicating a direction of a user's face in a real space which is detected by a motion sensor; a camera image data acquiring unit configured to acquire camera image data indicating the face direction with reference to the direction of the user's body which is specified based on a camera image; and a face direction specifying unit configured to specify the direction of the user's face based on the posture data, the motion sensor data, and the camera image data.

Abstract: In various exemplary embodiments a method for determining the waviness index of hair is provided. The method may comprise the steps of determining and/or establishing at least one hair examination region in at least one digital image, in which hair is depicted, determining a plurality of different brightness values in the hair examination region, and determining the waviness index of hair on the basis of the plurality of brightness values.

Abstract: The present invention provides a method for a transmission apparatus in a multimedia system. The method comprises a step of transmitting a message including a priority order of media source location information, and three or more pieces of media source location information, wherein the three or more pieces of media source location information include primary media source location information, alternative media source location information and original media source location information.

Abstract: The present invention is related to an image stitching method. It is further related to an image stitching device and to a computer readable medium carrying instructions for performing such method. The method according to the invention includes determining a pixel value of a pixel at an interpolation position in the region to be stitched using pixel values of corresponding pixels in a plurality of sub-images. According to the invention, the interpolation position corresponds to a weighted sum of positions of the corresponding pixels. The weighting factor for the position of a corresponding pixel in a given sub-image depends on a relative distance of at said interpolation position to a border of the region to be stitched associated with that sub-image.

Abstract: Certainty of detecting watermarks embedded in a color image is increased. Image date acquisition means of an image processing device acquires image data on a color image having a plurality of color channels. Numerical value acquisition means acquires respective numerical values in the plurality of color channels contained in a pixel value of the color image based on the image data acquired by the image data acquisition means. Image processing means embed a watermark in each of the plurality of color channels by changing the numerical value acquired by the numerical value acquisition means.

Abstract: The invention provides a device (10) for acquiring images of a plant, which device comprises: i) a first structure (60) defining a cavity (33) for receiving a portion to be imaged of the plant, the cavity presenting a longitudinal axis (13) and an opening (65) at one of its ends, the cavity being defined by at least one transparent wall forming a window (12) for observing the portion to be imaged; ii) a second structure (61) extending around the window (12) and mounted to move in translation along the longitudinal axis (13) relative to the first structure, and comprising: hollow reflective optics (11, 19) extending around the window and including a reflecting surface (19) directed towards the longitudinal axis; and a plurality of image sensors (18) regularly distributed around the window (12), the field of view (77) of each sensor including at least a portion of the reflective optics so as to capture images of the portion of the plant that is surrounded by the window, which images are reflected by the refl

Abstract: An artificial intelligence air conditioner may include a camera to obtain an image, an area recognition module for recognizing an area in which an occupant is located in an indoor space divided into a plurality of areas from the image obtained by the camera and distinguishing a living area with respect to the plurality of areas based on a result of recognizing a location of the occupant, and an airflow controller controlling airflow based on the distinguished living area, whereby airflow optimized for each living area of the indoor space may be controlled.

Abstract: A system processes data signals consisting of sums of independent signal terms, zero or more of which signal terms may already have been identified, in order to generate one or more additional terms. Deflated versions of the data signals are created by subtracting from the data signals any previously identified signal terms. Additional independent signal terms are computed using a set of reference signals organized into mutually independent partioning support sets. The images of each support set are computed on the data signals. Computed images on a data signal that are non-zero are identified as additional independent signal terms of that data signal.

Abstract: The unique watermark system comprising: identifying a presenter attendee, a first recipient attendee, and a second recipient attendee through a video conferencing session; detecting the presenter attendee sharing the visual content with the first recipient attendee and the second recipient attendee; selecting a first unique watermark and a second unique watermark from the plurality of unique watermarks and assigning them to a first recipient attendee and a second recipient attendee, respectively; inserting the first unique watermark into the visual content for the first recipient attendee and the second unique watermark into the visual content for the second recipient attendee; and transmitting the visual content with the first unique watermark to the first recipient attendee and the visual content with the second unique watermark to the second recipient attendee, wherein the method is performed by one or more special-purpose computing devices for hosting the video conferencing session.

Abstract: Disclosed are systems and methods for completing a transaction using a shelf and interactive tablet. The systems and methods may include receiving, at an interactive tablet comprising a processor, a captured image associated with an item; identifying, by the interactive tablet, a known image of the item based on the captured image of the item; and retrieving, by the interactive tablet, product information associated with the item based on the known image of the item.

Abstract: Methods, apparatus, systems and articles of manufacture are disclosed selectively generating and storing hashed reference signatures. An example method disclosed herein determining whether watermark coverage of a first media segment satisfies a dropout constraint, the first media segment corresponding to a first monitoring time interval of a media source feed and, when the watermark coverage of the first media segment does not satisfy the dropout constraint hashing first media signatures associated with the first media segment to generate corresponding first hashed signatures and generating first reference data for the first media segment, the first reference data including the first hashed signatures and the first media signatures. The example method further includes, when the watermark coverage of the first media segment satisfies the dropout constraint, generating second reference data for the first media segment, the second reference data including the first media signatures.

Abstract: Methods, systems, and apparatus, including steps of obtaining the source image having a plurality of modules, where a module corresponds to a smallest element that can be encoded in the source image; obtaining input data to be encoded; and encoding the input data in one or more modules of the plurality of modules of the source image. For each respective module of the one or more modules, the encoding includes determining, based on a corresponding value of the input data to be encoded in the respective module, a particular intensity level of multiple intensity levels associated with a particular color, where a number of the multiple intensity levels is greater than two; and encoding the particular color having the particular intensity level in the respective module of the source image.

Abstract: An image forming apparatus includes: an image forming device forming an image indicated by print data on a recording sheet; and a control device including a processor and functioning, through the processor executing a control program, as a controller configured to: extract text contained in the print data to determine a kind of a language of the text extracted; determine, for each determined kind of the language, whether to record the text consisting of the language of the determined kind; and when determining that the text is to be recorded, cause the image forming device to image form the text consisting of the language of the determined kind on the recording sheet, and, when determining that the text is not to be recorded, not to cause the image forming device to image form the text consisting of the language of the determined kind on the recording sheet.

Abstract: The invention relates to an optical measurement system (10) and measurement method for coaching an individual. The optical measurement system comprises a cloud database (8), a wireless communication device (7) and a vertical self-supporting equipment box (2) comprising a data processing device (20) comprising a computer program configured to measure ground contact and/or reaction time and position of a body part (1a) of a person (1) within a predetermined reaction area (6).

Abstract: A method for modelling age-related traits of a face, from a picture of the face is provided, wherein the age-related traits are either wrinkles or age spots, the method including: for each age-related trait of the face of the same nature, generating a vector including parameters of shape and appearance of the trait; and generating, from the generated vectors, a single representation vector modeling the age-related traits of the same nature in the face. The single representation vector stores information regarding the number of traits in the face and joint probabilities, over the face, of the shape and appearance features of the traits.

Abstract: Methods, systems, and devices that support determining whether media data has been altered are described. Captured media data may be segmented into one or more subsets, and cryptographic representations (e.g., hashes) based on the subsets may be written to an immutable ledger, possibly along with metadata and other related data. A block of a blockchain may be created for each entry in the immutable ledger. A set of media data may be validated, if a corresponding immutable ledger exists, based on segmenting the set of media data into one or more subsets in accordance with the segmenting upon capture, creating candidate cryptographic representations (e.g., hashes) based on the subsets, and comparing the candidate cryptographic representations with contents of the corresponding immutable ledger.

Abstract: Systems and methods of mobile database management and sharing are provided which include at least one mobile database and a first electronic circuit in communication with the mobile database. The system segments the mobile database into a plurality of database segments and indexes the plurality of database segments. The system sends one or more of the plurality of database segments to a second electronic circuit over a private secured communication protocol. One or both of the first and second electronic circuit may be housed within one or more mobile devices. Disclosed systems and methods include processing a database sharing request received from at least one mobile device wherein the database sharing request requests sharing of one or more of the plurality of database segments. A mobile device containing a database management and sharing system is also provided which includes a shared mobile database and one or more proxy entities related to the shared mobile database.

Abstract: A technique for managing parking of vehicles at a lower cost is disclosed. A parking management system includes a terminal device and a server device. The terminal device includes: a storage unit that stores user information; a reading unit that reads identification information of a parking space from a code associated with the parking space, and a transmission unit that transmits, to the server device, parking position information which includes the stored user information and the read identification information. The server device includes: a receiving unit that receives the parking position information from the terminal device; and an identifying unit that identifies a position where a vehicle is parked and a user of the parked vehicle, based on the received parking position information.

Abstract: A method for processing an acquired image (IA) showing a structured personal document (D) including a visual inspection zone (VIZ). The method includes determination of a first character string from an automatic reading zone or of barcode type (MRZ) of predetermined format, the first character string representing personalised information specific to the owner of the document (D), generation from the first character string of an intermediate image (IS) showing a second character string representing the same personalised information, detection in the acquired image of a third character string in the visual inspection zone (VIZ), representing the same personalised information, the detection comprising an overall comparison of the intermediate image (IS) with the acquired image, location in the visual inspection zone (VIZ) of a region of interest containing the third character string.

Abstract: Functionality is described for creating objects that represent respective items provided in a store. The objects can be embedded in various hosts, such as network-accessible pages (e.g., web pages), network-accessible journals (e.g., blogs), messages (e.g., Emails), and so on. Consumers can activate the objects in the hosts to retrieve information associated with the corresponding items. According to one illustrative feature, the functionality provides a general interface for creating objects for different types of items stored in the store, including file items of various kinds (e.g., text files, image files, audio files, spreadsheet files, video files, etc.), list items, folder items, and so on. According to another illustrative feature, the functionality provides an interface for allowing users to customize various aspects of objects.

Abstract: In one example, a smart mirror. The smart mirror includes a display adjacent one side of a half-mirror, sensors to acquire biometric information of a user facing an opposite side of the half-mirror, and a controller coupled to the sensors and the display. The controller determines from the biometric information a focal distance of the user. If the focal distance corresponds to the display, the controller presents a user interface on the smart mirror in a highly-visible manner and a reflection of the user in a less-visible manner. If the focal distance corresponds to the reflection of the user, the controller presents the user interface on the smart mirror in a less-visible manner and the reflection of the user in a highly-visible manner.

Abstract: A system includes a biometric sensor that detects biometric data of a person in a first region and a presuming unit that presumes that the person is a person to be tracked based on the biometric data. The system also includes a first camera acquiring first image data of the person in the first region and a second camera acquiring second image data of the person in a second region. The system further includes a monitoring unit that tracks the target person in the second region by checking the second image data with the first image data corresponding to the target person.

Abstract: Digital asset association techniques with search query data are described. In one example, A first digital asset is displayed at a location within digital content in a user interface. The location is specified using placeholder data of the digital content. An input is received via selection of an option as part of the user interface to initiate a search. Search query data associated with the placeholder data is obtained in response to the input. A search is initiated for a second digital asset using the search query data. The second digital asset is displayed at the location within the digital content in the user interface as specified by the placeholder data.

Abstract: Systems, and method and computer readable media that store instructions for scale invariant object detection.

Abstract: Various arrangements for identifying and grading cancer in tissue samples are presented. A digital image of a stained tissue sample may be acquired. A Shearlet transform may be performed on the digital image of the stained tissue sample. Shearlet coefficients may be calculated based on the performed Shearlet transform of the normalized digital RGB image of the stained tissue sample. A trained neural network may be applied to create a plurality of feature maps using the digital image and Shearlet coefficients, wherein the trained neural network was trained using a plurality of images and Shearlet coefficients of a plurality of digital images. A classifier may be applied to an output of the trained neural network to identify whether cancer is present in the stained tissue sample. A notification may be output that is indicative of a grade of detected cancer in the sample.

Abstract: A system and method for three-dimensional (3D) object detection is disclosed. A particular embodiment can be configured to: receive image data from at least one camera associated with an autonomous vehicle, the image data representing at least one image frame; use a trained deep learning module to determine pixel coordinates of a two-dimensional (2D) bounding box around an object detected in the image frame; use the trained deep learning module to determine vertices of a three-dimensional (3D) bounding box around the object; use a fitting module to obtain geological information related to a particular environment associated with the image frame and to obtain camera calibration information associated with the at least one camera; and use the fitting module to determine 3D attributes of the object using the 3D bounding box, the geological information, and the camera calibration information.

Abstract: High-precision local search is performed on the Internet. A map image-rendering software provider embeds spatial keys into maps, which are then provided to producers of Internet content such as map providers. For example, a homeowner may post a message on a web bulletin board advertising his house for sale, and including a map showing the location of the house. When a search engine's web crawler encounters a page having a spatial key embedded in an image, the spatial key is indexed with the other content on the page. Because the spatial key identifies a small geographic area, indexing the content with the spatial key allows search queries to be limited by area and still provide useful results. Thus, a user of a search engine searching for “house for sale” in a specific area will be directed to web pages that meet the geographic and content search terms.

Abstract: A measuring device for measuring an object. The measuring device includes a distance measuring unit having a beam emission unit generating a measurement radiation having a defined wavelength spectrum and a detector detecting the measurement radiation reflected on the object surface, the detector having at least one first sensor. The distance measuring unit generates distance measured data and reflection measured data as first distance measured data by emitting the measurement radiation and detecting the reflected measurement radiation according to the principle of triangulation. The measurement radiation is generated using a wavelength spectrum such that a fluorescence is excitable by an interaction of the measurement radiation with the object material to emit fluorescent light, wherein a spectrum of the fluorescent light and the wavelength spectrum of the measurement radiation are different.

Abstract: Some examples include receiving audio content through a microphone of an electronic device and determining whether embedded data is included in the received audio content. The electronic device may decode the received audio content to extract the embedded data. In addition, the electronic device may perform at least one of: sending a communication to a computing device over a network based on the extracted embedded data, or presenting information on a display of the electronic device based on the extracted embedded data.

Abstract: A damage assessment module operating on a computer system automatically evaluates a roof, estimating damage to the roof by analyzing a point cloud of a roof. The damage assessment module identifies individual shingles from the point cloud and detects potentially damaged areas on each of the shingles. The damage assessment module then maps the potentially damaged areas of each shingle back to the point cloud to determine which areas of the roof are damaged. Based on the estimation, the damage assessment module generates a report on the roof damage.

Abstract: A system, method, and apparatus for media content marking and tracking are disclosed.

Abstract: A person's smartphone serves as a security token—used variously to establish the user's authorization, and to test the bona fides of the counter-party system. In one illustrative ATM arrangement, a bank's computer system and the user's smartphone share secret information across a network. The smartphone employs this shared secret to generate a corresponding image, which is displayed on its screen. The displayed image is sensed by a camera in the ATM, which sends corresponding data back to the bank computer system. The bank computer system analyzes the received data and responds by sending information to the ATM authorizing a transaction. By such arrangement, authentication of the transaction involves one-way optical communication of visual information between the ATM and the smartphone, and two-way network communication of information between the ATM and the bank computer system. A great number of other features and arrangements are also detailed.

Abstract: The systems, methods, and computer program products disclosed herein pertain to image analysis techniques that may exploit computer vision, supervised and unsupervised machine learning, and deep neural network models in order to detect change and determine spatio-temporal attention regions of relevant change in multi-spectral and multi-resolution satellite images. In some embodiments, the automatic detection of changes in regions over large scales of satellite imagery data may be further used to generate alerts in the context of a monitoring and alerting system. In some embodiments, the input to the system may be provided through an analytics web interface or an application programming interface (API), and the output may also be served through the web interface or API.

Abstract: A system for monitoring the movement of people through a passageway comprising two or more time of flight sensors deployed across the passageway in a direction substantially transverse to the direction of movement of a person or object through the passageway. The system further comprises means to determine data regarding people or objects passing through the passageway based on the time of flight of a signal measured by the two or more sensors.