Abstract: A method and device are provided for detecting non-visible content of an object in a non-contact manner. The device includes a light source configured to emit light toward a surface of an object over a period of time. The device also includes an optical sensing component configured to receive a pattern of light from the surface of the object, and to record the received pattern at plural points in time. In addition, the device includes a processing component configured to determine temporal changes in the pattern during the plural points in time, and to detect whether motion is present in the object based on determined temporal changes in the pattern, where the motion represents a frequency source of non-visible content in the object.

Abstract: An interest degree determination device detects a gaze position of a first user included in a plurality of users on the basis of image information. When a second user other than the first user is present in the vicinity of the first user, the interest degree determination device determines whether the gaze position of the first user has been affected by the second user. The interest degree determination device stores a determination result of a determination unit and the gaze position of the first user in a storage unit so as to be associated with each other.

Abstract: Many conventional video processing algorithms attempting to detect human presence in a video stream often generate false positives on non-human movements such as plants moving in the wind, rotating fan, etc. To reduce false positives, a technique exploiting temporal correlation of non-human movements can accurately detect human occupancy while reject non-human movements. Specifically, the technique involves performing temporal analysis on a time-series signal generated based on an accumulation of foreground maps and an accumulation of motion map and analyzing the running mean and the running variance of the time-series signal. By determining whether the time-series signal is correlated in time, the technique is able to distinguish human movements and non-human movements. Besides having superior accuracy, the technique lends itself to an efficient algorithm which can be implemented on low cost, low power digital signal processor or other suitable hardware.

Abstract: Methods, systems, and computer program products are provided for the recognition of input of multiple objects into a computing device, wherein the computing device has a processor and at least one application for recognizing the input under control of the processor. The application is configured to determine at least one geometrical feature of a plurality of elements of the input, and compare the determined at least one geometrical feature with at least one pre-determined geometrical threshold to determine a positive or negative result. If the comparison yields a negative result, the application considers the elements as belonging to one object in the recognition of the input. If the comparison yields a positive result, the application considers the elements as belonging to multiple objects in the recognition of the input.

Abstract: A wearable retina/iris authentication system includes: a frame; a retina/iris scanner mounted to the frame for capturing images of a retina and/or an iris of a user; a mobile computing device, the mobile computing device including a memory for storing information about the user; a position determining device coupled to the mobile computing device for continuously determining a location of the user and providing the location to the mobile computing device; a display mounted to the frame and coupled to the mobile computing device for displaying information received from the mobile computing device; a database for storing information about one or more of a plurality of retina images and a plurality of iris images for matching with the captured images of one or more of the retina and the iris of the user; and a communication interface for comminuting with external systems remote to the wearable system.

Abstract: A storyboard interface displaying key frames of a video may be presented to a user. Individual key frames may represent individual shots of the video. Shots may be grouped based on similarity. Key frames may be displayed in a chronological order of the corresponding shots. Key frames of grouped shots may be spatially correlated within the storyboard interface. For example, shots of a common group may be spatially correlated so that they may be easily discernable as a group even though the shots may not be temporally consecutive and/or or even temporally close to each other in the timeframe of the video itself.

Abstract: A monitoring method for a region includes capturing an image frame of a monitored region using an image capturing device, analyzing the image frame to obtain the image information of one or more target subject, receiving one or more wireless signal from one or more wireless transmitter by a wireless receiver, determining whether to generate a warning signal according to one or more identification (ID) code and the image information. The wireless signal includes the ID code, and for the cases of multiple wireless signals, each wireless signal has a different ID code.

Abstract: A system for tracking a position of a cable attached between a power source and a machine in a worksite is disclosed. The cable is further coupled to a guiding device located in the worksite between the power source and the machine. The system includes a perception module disposed on the machine to detect a location of the guiding device. The system further includes a controller disposed to be in communication with the perception module. The controller determines the location of the guiding device with respect to a location of the machine and determines a distance between the guiding device and the machine. The controller further compares the distance with a first threshold distance and a second threshold distance and generates a warning if the distance is greater than the first threshold distance or smaller than the second threshold distance.

Abstract: An image processing apparatus synthesize a synthesized image by taking pixel values of pixels in a synthesized image corresponding to each point on the three dimensional projection plane as viewed from a specific viewpoint position, as pixel values of corresponding pixels of the first image based on the first correspondence relationship, and taking pixel values of each pixel in the synthesized image corresponding to pixels identified in the first image as being pixels representing a solid object, as pixel values of corresponding pixels of the second image based on the second correspondence relationship.

Abstract: A self-position calculating apparatus includes: a light projector 11 configured to project a patterned light beam 32a onto a road surface 31 around a vehicle; and a camera 12 configured to capture an image 38 of the road surface 31 around the vehicle including an area onto which the patterned light beam 32a is projected. The self-position calculating apparatus calculates an orientation angle of the vehicle 10 relative to the road surface 31 from a position of the patterned light beam 32a on the image 38 obtained by the camera 12, and calculates an amount of change in the orientation of the vehicle based on temporal changes in multiple feature points on the road surface which are detected from the image 38. The self-position calculating apparatus calculates current position and orientation angle of the vehicle by adding the amount of change in the orientation to initial position and orientation angle of the vehicle.

Abstract: A self-location calculating device projects a patterned light onto a road surface around a vehicle, as well as thereby captures and obtains an image of the road surface around the vehicle. Furthermore, the self-location calculating device detects a running condition of the vehicle. When it is determined that the vehicle is a running stable state, the self-location calculating device calculates a current position and a current attitude angle of the vehicle by adding an amount of change in an attitude to a predetermined initial attitude angle of the vehicle.

Abstract: An image acquiring system (100) includes an image acquiring apparatus (101) and a terminal (111). The image acquiring apparatus (101) acquires an image recorded by the terminal (111) and position information indicating a position at which the image is recorded, determines the state of a road included in the image based on the image, and associates and stores the state of the road and the position information. If the terminal (111) transmits to the image acquiring apparatus (101), a request for information concerning the state of a given road, the image acquiring apparatus (101) transmits the determined state of the road and the position information to the terminal (111).

Abstract: This malfunction diagnosis apparatus detects a line on a road surface by distinguishing the line from an asphalt surface other than the line. The malfunction diagnosis apparatus sets a first area corresponding to the line and a second area corresponding to the asphalt surface using a first image captured while the line is detected. The malfunction diagnosis apparatus calculates a first brightness in the first area and a second brightness in the second area using a second image captured while the vehicle is cruising and the line is not detected. Further, the malfunction diagnosis apparatus diagnoses the malfunction based on at least one of the first brightness and the second brightness.

Abstract: A system and process of nearsighted (myopia) camera object detection involves detecting the objects through edge detection and outlining or thickening them with a heavy border. Thickening may include making the object bold in the case of text characters. The bold characters are then much more apparent and heaver weighted than the background. Thresholding operations are then applied (usually multiple times) to the grayscale image to remove all but the darkest foreground objects in the background resulting in a nearsighted (myopic) image. Additional processes may be applied to the nearsighted image, such as morphological closing, contour tracing and bounding of the objects or characters. The bound objects or characters can then be averaged to provide repositioning feedback for the camera user. Processed images can then be captured and subjected to OCR to extract relevant information from the image.

Abstract: Methods and systems recognize alphanumeric characters in an image by computing individual representations of every character of an alphabet at every character position within a certain word transcription length. These methods and systems embed the individual representations of each alphabet character in a common vectorial subspace (using a matrix) and embed a received image of an alphanumeric word into the common vectorial subspace (using the matrix). Such methods and systems compute the utility value of the embedded alphabet characters at every one of the character positions with respect to the embedded alphanumeric character image; and compute the best transcription alphabet character of every one of the image characters based on the utility value of each embedded alphabet character at each character position. Such methods and systems then assign the best transcription alphabet character for each of the character positions to produce a recognized alphanumeric word within the received image.

Abstract: In one example a method of deceiving an optical augmentation device includes receiving an interrogation beam at an optical system from the optical augmentation device, encoding the interrogation beam with a false signature mis-identifying the optical system, and retro-reflecting the encoded interrogation beam to the optical augmentation device, without retro-reflecting the original interrogation beam.

Abstract: A system and process of nearsighted (myopia) camera object detection involves detecting the objects through edge detection and outlining or thickening them with a heavy border. Thickening may include making the object bold in the case of text characters. The bold characters are then much more apparent and heavier weighted than the background. Thresholding operations are then applied (usually multiple times) to the grayscale image to remove all but the darkest foreground objects in the background resulting in a nearsighted (myopic) image. Additional processes may be applied to the nearsighted image, such as morphological closing, contour tracing and bounding of the objects or characters. The bound objects or characters can then be averaged to provide repositioning feedback for the camera user. Processed images can then be captured and subjected to OCR to extract relevant information from the image.

Abstract: In one general embodiment, an article of manufacture includes an objective reference having at least two optical references. The optical references are selected from a group consisting of: a non-human-visible mark, a ruler, a spaced grid, a color calibration area, an area of reflectivity, a texture, and a pattern. In at other general embodiment, a method includes receiving an image of a product and an objective reference having at least two optical references. The optical references are selected from a group consisting of: a non-human-visible mark, a ruler, a spaced grid, a color calibration area, an area of reflectivity, a texture, and a pattern. The product is evaluated by comparing the product in the image to the optical references in the image.

Abstract: A feature amount generation device includes a feature point detecting unit which detects a feature point from an input image, a pixel information generating unit which generates pixel information from pixel values of the surroundings of the feature point in the input image, a feature point pixel information distribution generating unit which generates a first distribution of pixel information of a first region including the feature point, a sub-feature point detecting unit which detects sub-feature points that are feature points in the surroundings of the feature point, a sub-feature point pixel information distribution generating unit which generates second distributions of pixel information of second regions including the sub-feature points, and a feature amount generating unit which generates a feature amount at the feature point based on the first distribution mid the second distributions.

Abstract: A sensor data processing system and method is described. Contemplated systems and methods derive a first recognition trait of an object from a first data set that represents the object in a first environmental state. A second recognition trait of the object is then derived from a second data set that represents the object in a second environmental state. The sensor data processing systems and methods then identifies a mapping of elements of the first and second recognition traits in a new representation space. The mapping of elements satisfies a variance criterion for corresponding elements, which allows the mapping to be used for object recognition. The sensor data processing systems and methods described herein provide new object recognition techniques that are computationally efficient and can be performed in real-time by the mobile phone technology that is currently available.

Abstract: A system and method enable semantic comparisons to be made between word images and concepts. Training word images and their concept labels are used to learn parameters of a neural network for embedding word images and concepts in a semantic subspace in which comparisons can be made between word images and concepts without the need for transcribing the text content of the word image. The training of the neural network aims to minimize a ranking loss over the training set where non relevant concepts for an image which are ranked more highly than relevant ones penalize the ranking loss.

Abstract: A method for extracting a representation from an image includes inputting an image to a pre-trained neural network. The gradient of a loss function is computed with respect to parameters of the neural network, for the image. A gradient representation is extracted for the image based on the computed gradients, which can be used, for example, for classification or retrieval.

Abstract: Context-based priors are utilized in machine learning networks (e.g., neural networks) for detecting objects in images. The likely locations of objects are estimated based on context labels. A machine learning network identifies a context label of an entire image. Based on the, the network selects a set of likely regions for detecting objects of interest in the image.

Abstract: A light-source-driving control device thins a pixel at an edge portion of image data by a first pixel width; sets a second pixel width at the edge portion such that a ratio between the first and second pixel widths is constant; creates light amount information including a first light amount with which a pixel not included in the set second pixel width is exposed, and a second light amount with which a pixel included in the second pixel width is exposed; selects the first or second light amount based on a power-modulation control signal; drives light sources according to the selected light amount and a light-source-modulation pulse signal; and counts an addition value of pixel density considering a pixel width and a light amount, based on a data-width setting signal based on the second pixel width, and on a light-amount setting signal based on the light amount information.

Abstract: An image processing apparatus which prevents a printing process from taking long. The image processing apparatus is connected to a printing apparatus having a storage unit and transfers image data to the printing apparatus. A size of predetermined image data when a speed at which the predetermined image data is stored in the storage unit and a speed at which the predetermined image data stored in the storage unit is printed are the same is obtained as a threshold size. A size of the image data stored in the storage unit in a compressed form is estimated. A compression method for use in compressing the transferred image data is selected from a plurality of compression methods based on the threshold size and the estimated size of the image data. The image processing apparatus notifies the printing apparatus of the determined compression method.

Abstract: A printing control apparatus includes a filter processing portion configured to execute a first filter process of inserting a first description representing a content of processing by a first filter into description data described in a prescribed format for representing an image, and execute a second filter process of inserting a second description representing a content of processing by a second filter into the description data; and a printing data generating portion configured to interpret and convert the description data into printing data. Further, in the first filter process, the filter processing portion inserts the first description into the description data together with identification information associated with the first filter, and in the second filter process, the filter processing portion determines a position of the insertion of the second description within the description data on the basis of the identification information, and inserts the second description at the determined position.

Abstract: A pharmaceutical filling system for a high volume pharmacy is described. The system can include a paper feed device and method. The system may include a paper collator with at one selectively openable shelf. A control device may monitor the capacity of the selectively openable shelf, and instruct a printer to hold printing additional print jobs. The selectively openable shelf may then be opened to allow the printed literature resting thereon to fall to a lower shelf, and the printer can then be restarted. Sheets of printed literature may be drawn from the paper collator, collected with any other sheets of printed literature associated with a prescription order, and mated with the prescription order for shipment.

Abstract: Provided are a label printing control device, a non-transitory computer-readable storage medium storing a label printing control program and a label printing control method. The label printing control device includes a cut mark extraction section that obtains print data including at least label images and corresponding cut marks and extracts the cut marks from the print data, and an end product image creation section that creates an end product image by removing the cut marks from the print data. The label printing control device further includes an apex finding section that finds apexes of the cut marks, an adjusting image creation section that creates an adjusting image including the cut marks with graduations added adjacent to each apex, and a print instruction section that instructs a label printing device to print the end product image and the adjusting image successively on pre-die-cut label material.

Abstract: The present invention concerns a RFID transponder, such as a card, comprising a first chip electrically connected to a first antenna and a second chip electrically connected to a second antenna. The first antenna comprises a secondary antenna which is inductively to the second antenna.

Abstract: Implementing a neural network can include receiving a macro instruction for implementing the neural network within a control unit of a neural network processor. The macro instruction can indicate a first data set, a second data set, a macro operation for the neural network, and a mode of operation for performing the macro operation. The macro operation can be automatically initiated using a processing unit of the neural network processor by applying the second data set to the first data set based on the mode of operation.

Abstract: In some examples, a machine learning system may use morphological knowledge to enhance a deep learning framework for learning word embedding. The system may consider, among other things, morphological similarities between and among words in a learning process so as to handle new or rare words, edit distances, longest common substring similarities, morpheme similarities, and syllable similarities as morphological knowledge to build a relation matrix between or among words. The system may apply the deep learning framework to query classification, web search, text mining, information retrieval, and natural language processing tasks, for example. The system may accomplish such tasks with relatively high efficiency and speed, while utilizing less computing resources as compared to other systems.

Abstract: An analog neuromorphic circuit is disclosed having resistive memories that provide a resistance to an input voltage signal as the input voltage signal propagates through the resistive memories generating a first output voltage signal and to provide a resistance to a first error signal that propagates through the resistive memories generating a second output voltage signal. A comparator generates the first error signal that is representative of a difference between the first output voltage signal and the desired output signal and generates the first error signal so that the first error signal propagates back through the plurality of resistive memories. A resistance adjuster adjusts a resistance value associated with each resistive memory based on the first error signal and the second output voltage signal to decrease the difference between the first output voltage signal and the desired output signal.

Abstract: Embodiments of the present invention relate to finding semantic parts in images. In implementation, a convolutional neural network (CNN) is applied to a set of images to extract features for each image. Each feature is defined by a feature vector that enables a subset of the set of images to be clustered in accordance with a similarity between feature vectors. Normalized cuts may be utilized to help preserve pose within each cluster. The images in the cluster are aligned and part proposals are generated by sampling various regions in various sizes across the aligned images. To determine which part proposal corresponds to a semantic part, a classifier is trained for each part proposal and semantic part to determine which part proposal best fits the correlation pattern given by the true semantic part. In this way, semantic parts in images can be identified without any previous part annotations.

Abstract: Systems and methods are provided to engage in multi-tiered optimization where there may be a first multi-objective optimization and a second constraint optimization. The multi-objective optimization may be used to drive to one or more goals of the optimization problem. The constraint optimization or minimization may be used to drive towards a reduced and/or no constraint situation where the solution to the overall problem is feasible or near-feasible.

Abstract: A distributable and serializable finite state machine and methods for using the distributable and serializable finite state machine are provided wherein finite state machine instance can be location-shifted, time-shifted or location-shift and time-shifted, for example by serializing and deserializing each instance. Each instance can be located-shifted between agents, and a persistent memory storage location is provided to facilitate both location-shifting and time-shifting. Finite state machine instances and the actions that make up each instance can be run in a distributed fashion among a plurality of agents.

Abstract: A method for object classification in a decision tree based adaptive boosting (AdaBoost) classifier implemented on a single-instruction multiple-data (SIMD) processor is provided that includes receiving feature vectors extracted from N consecutive window positions in an image in a memory coupled to the SIMD processor and evaluating the N consecutive window positions concurrently by the AdaBoost classifier using the feature vectors and vector instructions of the SIMD processor, in which the AdaBoost classifier concurrently traverses decision trees for the N consecutive window positions until classification is complete for the N consecutive window positions.

Abstract: A cultural determination engine is provided. The cultural determination engine is configured to assess a cultural profile by computing a distance between a received response to a query and a reference response.

Abstract: A method estimates similarities in data streams. A data source receives input vectors from users. A sketch feature generator converts the input vectors into sketch feature vectors corresponding to the users, wherein each sketch feature vector represents data and meta-information from each user received in a most recent sample period. A similarity comparator compares each sketch feature vector against other sketch feature vectors to calculate similarity probabilities between the users. A processor running a decision loop determines a prediction result for at least one user based on the similarity probabilities.

Abstract: A system can generate a heavy load pre-warning or an overload pre-warning for distribution transformers. Operation of the system can include selecting data records received from a plurality of data sources; converting the data records in the plurality of different data formats; filtering the data records in the database by using a predetermined threshold and matching each of the filtered data records with one of a plurality of distribution transformers; transforming the matched data records to a plurality of predefined predictor variables; selecting a subset of the plurality of predefined predictor variables; training, testing and tuning a model and forecasting at least one of heavy load or overload for each of the plurality of distribution transformers in a predetermined region.

Abstract: A computer implemented method and system for determining reliability of a machine includes receiving one of a machine data from one or more locations through an internet of things (IOT) based machine wearable sensor network. The method further includes storing the data in a distributed computer database communicatively coupled to an enterprise resource planning (ERP) system and extracting, through a computer server, one or more entity information from the data to compare against a pre-defined baseline. Further, mapping, though a big data machine learning engine, the extracted one or more entity information into a multi-classification model. The method includes indicating, through a machine learning engine coupled to a predictive analytics engine, on a user interface a set of analytical predictions for machine maintenance, repair and operation.

Abstract: Disclosed herein is a visual analytics system and method that provides a proactive and predictive environment in order to assist decision makers in making effective resource allocation and deployment decisions. The challenges involved with such predictive analytics processes include end-users' understanding, and the application of the underlying statistical algorithms at the right spatiotemporal granularity levels so that good prediction estimates can be established. In the disclosed approach, a suite of natural scale templates and methods are provided allowing users to focus and drill down to appropriate geospatial and temporal resolution levels. The disclosed forecasting technique is based on the Seasonal Trend decomposition based on Loess (STL) method applied in a spatiotemporal visual analytics context to provide analysts with predicted levels of future activity.

Abstract: A device management method of a device management apparatus that manages information of devices installed in a home, includes: acquiring electric power consumption per unit time in predetermined time intervals, for each of branch circuits of a power distribution board to which the devices can be connected; estimating existence of the devices connected to the branch circuit, based on change of electric power consumption over time, per unit time; condition determining whether or not an estimated device, which is a device regarding which existence has been estimated in the estimating, satisfies a predetermined condition; and transmitting inquiry information to a user in accordance with the predetermined condition, only in a case where determination is made in the condition determining that the estimated device satisfies the predetermined condition.

Abstract: In various example embodiments, a semantic modeling server includes a semantic model and an inductive logic programming module. The semantic module includes underlying data that defines one or more characteristics of a part to be manufactured. The inductive logic programming module is provided with positive and negative examples of a feature to be identified and part data that defines the part to be manufactured. Given the examples of the feature and the semantic model, the inductive logic programming module determines various rules that can be used to identify whether the provided part data includes the feature defined by the semantic model. Using the determined rules the inductive logic programming module then identifies instances of the feature associated with the semantic model within the provided part data. The inductive logic programming can then be iteratively executed with the semantic model to refine the determined rules.

Abstract: Described are methods, systems, and computer program products for correlating operational information with subsequent action information. Operational information is received from a computing system performing one or more services. Pattern matching is performed for the operational information with data within a knowledge base. Based on the results of the pattern matching, a subsequent action associated with the operational information is determined.

Abstract: Various embodiments include methods implemented on a mobile communication device for using a contact-based predictive response database on a communication device. The methods include determining communication context information from a communication exchange between a user using the communication device and a contact, selecting a predictive response database associated with the contact, determining a suggested communication element from the predictive response database based on a prior communication element identified from the communication exchange and the communication context information, and providing the suggested communication element to the user.

Abstract: An appliance inference device includes a receiver, a first storage, an analyzer, an inference module and a second storage. The first storage stores pieces of power information in each of which a value of electric power received by the receiver is correlated with a time. When a piece of power information stored in the first storage satisfies a prescribed condition, the analyzer classifies the piece of power information as a group corresponding to the condition. The inference module checks the piece of power information classified as the group by the analyzer with appliance information stored in the second storage, thereby inferring the name of an electrical appliance corresponding to the group. The second storage stores the appliance information, with the information correlated with the name of the electrical appliance.

Abstract: Systems and methods that employ interactions between quantum computing systems and digital computing systems are described. For an iterative method, a quantum computing system may be designed, operated, and/or adapted to provide a rate of convergence that is greater than the rate of convergence of a digital supercomputer. When the digital supercomputer is iteratively used to evaluate an objective function at a cost incurred of C per iteration, the quantum computing system may be used to provide the input parameter(s) to the objective function and quickly converge on the input parameter(s) that optimize the objective function. Thus, a quantum computing system may be used to minimize the total cost incurred CT for consumption of digital supercomputer resources when a digital supercomputer is iteratively employed to evaluate an objective function.

Abstract: Examples of the present disclosure describe systems and methods of transfer learning techniques for disparate label sets. In aspects, a data set may be accessed on a server device. The data set may comprise labels and word sets associated with the labels. The server device may induce label embedding within the data set. The embedded labels may be represented by multi-dimensional vectors that correspond to particular labels. The vectors may be used to construct label mappings for the data set. The label mappings may be used to train a model to perform domain adaptation or transfer learning techniques. The model may be used to provide results to a statement/query or to train a different model.

Abstract: Technologies are generally described for methods and systems in a machine learning environment. In an example, a method may include receiving training data and test data. The method may also include determining a set of out of sample errors based on the training data and test data. The method may also include determining a set of gradient values based on the out of sample errors. Each gradient value may relate to a first magnitude and to a second magnitude. The first magnitude may be associated with the set of out of sample errors. The second magnitude may be associated with out of sample errors of a portion of the training data excluding a particular point of the test data. The method may also include transforming the set of gradient values into modified density data. The method may also include generating alternative training data based on the modified density data.

Abstract: A method for receiving a description of a problem and applying machine learning to automatically solve the problem includes receiving, by a first computing device, from a second computing device, via a user interface component, a description of a first problem. The method includes assigning, by a clustering engine executing on the first computing device, the first problem to a class. The method includes identifying, by a correlation engine executing on the first computing device, a first database associated with the class. The method includes retrieving, by the correlation engine, from the identified database, first data relevant to the first problem. The method includes providing, by the first computing device, via the user interface component, a suggestion for solving the first problem, based on the retrieved first data.