Abstract: An attribute recognition apparatus including a unit for extracting a first feature from an image by using a feature extraction neural network; a unit for recognizing a first attribute of an object in the image based on the first feature by using a first recognition neural network; a unit for determining a second recognition neural network from a plurality of second recognition neural network candidates based on the first attribute; and a unit for recognizing at least one second attribute of the object based on the first feature by using a second recognition neural network.

Abstract: An image annotating method includes: acquiring an image collected at a terminal; acquiring voice information associated with the image; annotating the image according to the voice information; and storing an annotated result of the image.

Abstract: This disclosure relates to systems and methods for solving generic inverse problems by providing a coupled representation architecture using transform learning. Convention solutions are complex, require long training and testing times, reconstruction quality also may not be suitable for all applications. Furthermore, they preclude application to real-time scenarios due to the mentioned inherent lacunae. The methods provided herein require involve very low computational complexity with a need for only three matrix-vector products, and requires very short training and testing times, which makes it applicable for real-time applications. Unlike the conventional learning architectures using inductive approaches, the CASC of the present disclosure can learn directly from the source domain and the number of features in a source domain may not be necessarily equal to the number of features in a target domain.

Abstract: A cloud computing system can be configured to generate a synthetic data stream that tracks a reference data stream. A model optimizer of the cloud computing system can receive, from an interface of the cloud computing system, a synthetic data stream request indicating a reference data stream. A dataset generator of the cloud computing system can generate a synthetic data stream that tracks the reference data stream by repeatedly swapping data models of the reference data stream. One such repeat can include retrieving, by the dataset generator from a model storage, a current data model of the reference data stream and generating a new data model of the reference data stream. The model optimizer can store the new data model in the model storage. The dataset generator can generate a synthetic data stream using the current data model of the reference data stream.

Abstract: An exemplary system, method, and computer-accessible medium can include, for example, receiving an original dataset(s), receiving a synthetic dataset(s), training a model(s) using the original dataset(s) and the synthetic dataset(s), and evaluating the synthetic dataset(s) based on the training of the model(s). The model(s) can include a first model and a second model, and the first model can be trained using the original dataset(s) and the second model can be trained using the synthetic dataset(s). The synthetic dataset(s) can be evaluated by comparing first results from the training of the first model to second results from the training of the second model.

Abstract: A system for returning synthetic database query results. The system may include a memory unit for storing instructions, and a processor configured to execute the instructions to perform operations comprising: receiving a query input by a user at a user interface; determining, based on natural language processing, a type of the query input; determining, based on the received query input and a database language interpreter, an output data format; returning, based on a generation model and the output data format, a result of the query input; providing, to a plurality of training models and based on the determined query type, the query input and the result; and training the training models, based on the query input and the result.

Abstract: This learning data generation device (10) is provided with: an identification unit (11) which identifies a subject included in a first captured image, and generates an identification result in which information indicating the type and existence of the identified subject or the motion of the identified subject is associated with the first captured image; and a generation unit (12) which generates learning data on the basis of the identification result and a second captured image, which is associated with the first captured image but is different in type from the first captured image.

Abstract: An active learning system classifies multiple objects in an input image from the set of images with a classification metric indicative of uncertainty of each of the classified object to belong to one or different classes and determines a diversity metric of the input image indicative of diversity of the objects classified in the input image. The active learning system evaluates the diversity metric of the input image and to cause rendering of the input image on a display device based on a result of the evaluation and trains the classifier using the labelled objects of the input image.

Abstract: Systems and techniques for classification and localization based on annotation information are presented. In one example, a system trains a convolutional neural network based on training data and a plurality of images. The training data is associated with a plurality of patients from at least one imaging device. The plurality of images is associated with a plurality of masks from a plurality of objects. The convolutional neural network comprises a decoder consisting of at least one up-sampling layer and at least one convolutional layer. The system also generates a loss function based on the plurality of masks, where the loss function is iteratively back propagated to tune parameters of the convolutional neural network. The system also predicts a classification label for an input image based on the convolutional neural network.

Abstract: Provided is an apparatus for generating data based on generative adversarial networks (GANs), the apparatus including a first generator configured to receive input data and generate a first fake image and a first discriminator configured to receive the first fake image generated by the first generator and a first real image and verify whether an image is fake or real.

Abstract: A machine learning based recommendation model, including a supervised learning classifier configured to receive input training data that includes a plurality of behavioral determinants, a supervised learning model configured to receive subject input data that includes a plurality of behavior determinants, wherein the supervised learning model outputs a predicted behavior of a subject, and a channel selection module configured to receive the subject input data and the predicted behavior and to determine a recommended communication channel for the subject to follow to achieve the predicted behavior.

Abstract: Systems and techniques for classification based on annotation information are presented. In one example, a system trains a convolutional neural network based on training data and a plurality of images. The training data is associated with a plurality of patients from at least one imaging device. The plurality of images is associated with a plurality of masks from a plurality of objects. The system also generates a loss function based on the plurality of masks, where the loss function is iteratively back propagated to tune parameters of the convolutional neural network. The system also predicts a classification label for an input image based on the convolutional neural network.

Abstract: An image processing device has an extraction unit that extracts image data by using a predetermined sliding window in an original image; a learning unit that generates a prediction model by performing machine learning on learning data including the image data by using a teaching signal representing classification of the image data; and a select unit that selects, out of other image data different from the image data, another image data in which an error in classification based on the prediction model is larger than a predetermined threshold and adds the selected another image data to the learning data, and the learning unit updates the prediction model by repeating the machine learning in the learning data in which the another image data is added.

Abstract: A system and method for detecting data drift is disclosed. The system may be configured to perform a method, the method including receiving model training data and generating a predictive model. Generating the predictive model may include model training or hyperparameter tuning. The method may include receiving model input data and generating predicted data using the predictive model, based on the model input data. The method may include receiving event data and detecting data drift based on the predicted data and the event data. The method may include receiving current data and detecting data drift based on the data profile of the current data. The method may include model training and detecting data drift based on a difference in a trained model parameter from a baseline model parameter. The method may include hyperparameter tuning and detecting data drift based on a difference in a tuned hyperparameter from a baseline hyperparameter.

Abstract: An electronic apparatus for recognizing a user and a method therefor are provided. The electronic apparatus includes a communication interface, a dynamic vision sensor (DVS), a memory including a database in which one or more images are stored, and at least one processor. The at least one processor is configured to generate an image, in which a shape of an object is included, based on an event detected through the DVS, control the memory to store a plurality of images generated under a specified condition, in the database, identify shapes of the user included in each of the plurality of images stored in the database, and generate shape information for recognizing the user based on the identified shapes. The plurality of images may include a shape of a user.

Abstract: Systems and methods for generating synthetic data are disclosed. For example, a system may include one or more memory units storing instructions and one or more processors configured to execute the instructions to perform operations. The operations may include receiving a dataset including time-series data. The operations may include generating a plurality of data segments based on the dataset, determining respective segment parameters of the data segments, and determining respective distribution measures of the data segments. The operations may include training a parameter model to generate synthetic segment parameters. Training the parameter model may be based on the segment parameters. The operations may include training a distribution model to generate synthetic data segments. Training the distribution model may be based on the distribution measures and the segment parameters.

Abstract: A location of an object of interest (205) is determined using both observations and non-observations. Numerous images (341-345) are stored in a database in association with image capture information, including an image capture location (221-225). Image recognition is used to determine which of the images include the object of interest (205) and which of the images do not include the object of interest. For each of multiple candidate locations (455) within an area of the captured images, a likelihood value of the object of interest existing at the candidate location is calculated using the image capture information for images determined to include the object of interest and using the image capture information for images determined not to include the object of interest. The location of the object is determined using the likelihood values for the multiple candidate locations.

Abstract: Systems and techniques for classification based on annotation information are presented. In one example, a system trains a convolutional neural network based on training data and a plurality of images. The plurality of images is associated with a plurality of masks, a plurality of image level labels, and/or a bounding box. The system also generates a first loss function based on the plurality of masks, a second loss function based on the plurality of image level labels, and a third loss function based on the bounding box. Furthermore, the system generates a fourth loss function based on the first loss function, the second loss function and the third loss function, where the fourth loss function is iteratively back propagated to tune parameters of the convolutional neural network. The system also predicts a classification label for an input image based on the convolutional neural network.

Abstract: Systems and techniques are disclosed for labeling objects within an image. The objects may be labeled by selecting an option from a plurality of options such that each option is a potential label for the object. An option may have an option score associated with. Additionally, a relation score may be calculated for a first option and a second option corresponding to a second object in an image. The relation score may be based on a frequency, probability, or observance corresponding to the co-occurrence of text associated with the first option and the second option in a text corpus such as the World Wide Web. An option may be selected as a label for an object based on a global score calculated based at least on an option score and relation score associated with the option.

Abstract: Examples are disclosed herein that relate to entity tracking. One examples provides a computing device comprising a logic processor and a storage device holding instructions executable by the logic processor to receive image data of an environment including a person, process the image data using a face detection algorithm to produce a first face detection output at a first frequency, determine an identity of the person based on the first face detection output, and process the image data using another algorithm that uses less computational resources of the computing device than the face detection algorithm. The instructions are further executable to track the person within the environment based on the tracking output, and perform one or more of updating the other algorithm using a second face detection output, and updating the face detection algorithm using the tracking output.

Abstract: The present invention is a collation/retrieval system collating a product manufactured by or delivered from a producer or a distributor with a product to be collated comprising: a storage unit that stores an image feature of a predetermined collation area of the product determined in advance at a position relative to a reference section common in every product; a to-be-collated product feature extraction unit that receives an image of the product to be collated and detecting the reference section of the product from the received image to extract an image feature of the collation area determined by reference to the reference section; and a collation unit that collates the stored image feature with the image feature of the collation area of the product to be collated.

Abstract: According to an embodiment of this invention, an image processing apparatus for acquiring image data edited after obtained by image-capturing an object performs the following processing. More specifically, the apparatus obtains information, equivalent to a distance from a focal plane in image-capturing, corresponding to image data before editing, obtained when image-capturing the object, and generates information, equivalent to a distance from a focal plane, corresponding to the edited image data, based on the edited image data, and the information, equivalent to the distance from the focal plane, corresponding to the image data before editing.

Abstract: A method for improving the abrasion resistance of a printed image, the image being processed digitally by a colour management system for a print process comprising a set of colorants, the colour management system transforming a colour of a part of the image into a colorant coverage for said part of the printed image, wherein a sum of coverage values for all colorants in the set of colorants is larger than a user-selectable threshold value. In a preferred embodiment the set of colorants includes a colorless, transparent colorant.

Abstract: A control method for an information processing apparatus that displays an edit screen of print data for printing a predetermined page, comprising: displaying, on the edit screen, a print region that includes a binding position indicating a position at which binding is performed using a binding material in the predetermined page and indicates a printable region in the predetermined page; receiving designation of a layout position of an image in the predetermined page; and making a notification if the designation is received so that the image is laid out in a predetermined region including at least a partial region in the print region between the binding position and a side of the print region on a side of the binding position, wherein a printing apparatus executes printing based on the print data edited in the edit screen.

Abstract: Optical articles including a spatially defined arrangement of a plurality of data rich retroreflective elements, wherein the plurality of retroreflective elements comprise retroreflective elements having at least two different retroreflective properties and at least two different optical contrasts with respect to a background substrate when observed within an ultraviolet spectrum, a visible spectrum, a near-infrared spectrum, or a combination thereof.

Abstract: A wearable device including: a housing including a slot operable to contain an electronically readable tag or a transmitter; and a plurality of depressible buttons coupled to the housing, wherein depressing at least two of the plurality of depressible buttons in a predetermined order is operable to expose the electronically readable tag in or from the slot or is operable to activate the transmitter to transmit a signal. A method of exposing an electronically readable tag including; exposing an electronic tag contained in a housing of a wearable device by pressing a plurality of buttons coupled to the housing in a predetermined order; and exposing the electronic tag to a tag reader.

Abstract: A dynamic transaction card that is manufactured using conductive plastic jumpers that will dissolve when in contact with a solvent used to tamper with the dynamic transaction card. Internal components of a dynamic transaction card may be manufactured using a synthetic or semi-synthetic organic material, such as, for example, plastics. These materials may be conductive to provide functionality to a dynamic transaction card, such as a connection between an integrated circuit and other card components such that when the materials dissolve, the connections are broken and the dynamic transaction card may be inactive due to the loss of various connections.

Abstract: The invention relates to a label comprising an electronic chip for applying to a bottle of liquid, said label comprising; a first layer (S), called a support layer (S), a second layer (E), called an electronic chip layer (E), comprising at least an electronic chip and an antenna connected to the electronic chip, and a third layer (P), called customization layer (P), in which each layer (S, E, P) has two faces, and the three layers (S, E, P) are placed one above the other in a stacking direction, and in which the customization layer (P) comprises at least one sublayer (C5), called metal sublayer (C5), made of a metal material, the thickness of the metal sublayer (C5) along the stacking direction being less than or equal to 35 ?m.

Abstract: Materials, methods of making, and methods of using an integrated wireless detector for real time interrogating metallic tubular structures comprising: an RF patch antenna; a passive surface acoustic wave (SAW) sensor; and data analytic methodologies. An embodiment relates to interrogating a metallic structure having a uniform cross section using an antenna which launches electromagnetic radiation. A sensor may be located within the structure is configured to re-emit electromagnetic radiation modified depending on parameters for which the sensor has been functionalized. An antenna may receive radiation as modified by the sensor, or may receive the transmitted or backscattered radiation directly, without use of a sensor. The antenna then communicates wirelessly with an interrogator providing data which may be used to understand the operational status of the structure in real-time.

Abstract: Implementations are directed to methods for providing an enhanced encounter via a holographic virtual assistant, including detecting, by one or more processors, an encounter request from a user, selecting a first encounter including a first holographic representation and a first dialog output, providing the first encounter for presentation to the user on the holographic virtual assistant, receiving, from the user, a first user reaction, the first user reaction including a first user dialog input and a first user engagement input, and training, using the first user reaction, a conversational goal model.

Abstract: Systems and methods for transforming legacy models and transforming a model into a neural network model are disclosed. In an embodiment, a method may include receiving input data comprising an input model, an input dataset, and an input command. The method may include applying the input model to the input dataset to generate model output and storing model output and at least one of input model features or a map of the input model. The method may include generating a candidate neural network models with parameters. The method may include tuning the candidate neural network models to the input model. The method may include receiving model output from the candidate neural network models and selecting a neural network model from the candidate neural network models based on the candidate model output and the model selection criteria. In some aspects, the method may include returning the selected neural network model.

Abstract: Anomaly detection from time series is one of the key components in automated monitoring of one or more entities. Domain-driven sensor selection for anomaly detection is restricted by knowledge of important sensors to capture only a certain set of anomalies from the entire set of possible anomalies. Hence, existing anomaly detection approaches are not very effective for multi-dimensional time series. Embodiments of the present disclosure depict sparse neural network for anomaly detection in multi-dimensional time series (MDTS) corresponding to a plurality of parameters of entities. A reduced-dimensional time series is obtained from the MDTS via an at least one feedforward layer by using a dimensionality reduction model. The dimensionality reduction model and recurrent neural network (RNN) encoder-decoder model are simultaneously learned to obtain a multi-layered sparse neural network. A plurality of error vectors corresponding to at least one time instance of the MDTS is computed to obtain an anomaly score.

Abstract: Computer implemented methods and systems are disclosed for obtaining predictions of legal events, such as legal and factual arguments presented to courts, juries or other adjudicative or fact-finding bodies, using machine-learning algorithms, wherein (i) unstructured data, such as natural language text from documents, such as pleadings, briefs or corpuses of evidence are converted into tokens, vectors and/or embeddings; (ii) the machine-learning algorithm(s) are provided the converted unstructured data as inputs; and (iii) the machine-learning algorithms provide confidence or probability scores predicting outcomes of legal events, such as legal proceedings or one or more legal or factual issues to be decided by particular adjudicators, tribunals or fact-finding bodies.

Abstract: Disclosed is an apparatus for determining goodness of fit related to microphone placement capable of communicating with other electronic devices and an external server in a 5G communication network, in which an artificial intelligence (AI) algorithm and/or a machine learning algorithm are executed. The apparatus includes an inputter, a communicator, a storage, and a processor. As the apparatus is provided, sound recognition effects can be improved.

Abstract: Estimating Remaining Useful Life (RUL) from multi-sensor time series data is difficult through manual inspection. Current machine learning and data analytics methods, for RUL estimation require large number of failed instances for training, which are rarely available in practice, and these methods cannot use information from currently operational censored instances since their failure time is unknown. Embodiments of the present disclosure provide systems and methods for estimating RUL using time series data by implementing an LSTM-RNN based ordinal regression technique, wherein during training RUL value of failed instance(s) is encoded into a vector which is given as a target to the model. Unlike a failed instance, the exact RUL for a censored instance is unknown. For using the censored instances, target vectors are generated and the objective function is modified for training wherein the trained LSTM-RNN based ordinal regression is applied on an input test time series for RUL estimation.

Abstract: A joint position estimation device including a memory, and a processor connected to the memory. The processor executes a process including estimating, by a first DNN for which a first parameter determined by learning of the first DNN has been set, a body part region of the animal with respect to input image to be processed; and estimating, by the second DNN for which a second parameter determined by learning of the second DNN has been set, a first joint position and a second joint position in each of the body part region estimated by the first DNN and a plural body parts region in which a plurality of the body part regions are connected.

Abstract: A computer device for training a deep neural network is provided. The computer device includes a receiving unit for receiving a two-dimensional input image frame, and a deep neural network for examining the two-dimensional input image frame in view of objects being included in the two-dimensional input image frame. The deep neural network includes a plurality of hidden layers and an output layer representing a decision layer. The computer device includes a training unit for training the deep neural network using transfer learning based on synthetic images for generating a model comprising trained parameters, and an output unit for outputting a result of the deep neural network based on the model.

Abstract: Enhanced techniques and circuitry are presented herein for artificial neural networks. These artificial neural networks are formed from artificial neurons, which in the implementations herein comprise a memory array having non-volatile memory elements. Neural connections among the artificial neurons are formed by interconnect circuitry coupled to input control lines and output control lines of the memory array to subdivide the memory array into a plurality of layers of the artificial neural network. Control circuitry is configured to transmit a plurality of iterations of an input value on input control lines of a first layer of the artificial neural network for inference operations by at least one or more additional layers. The control circuitry is also configured to apply an averaging function across output values successively presented on output control lines of a last layer of the artificial neural network from each iteration of the input value.

Abstract: A neuromorphic system includes an address translation device that translates an address corresponding to each of synaptic weights between presynaptic neurons and postsynaptic neurons to generate a translation address, and a plurality of synapse memories that store the synaptic weights based on the translation address. The translation address is generated such that at least two of synaptic weights corresponding to each of the postsynaptic neurons are stored in different synapse memories of the plurality of synapse memories and such that at least two of synaptic weights corresponding to each of the presynaptic neurons are stored in different synapse memories.

Abstract: Provided is a learning device of a neural network including a bitwidth reducing unit, a learning unit, and a memory. The bitwidth reducing unit executes a first quantization that applies a first quantization area to a numerical value to be calculated in a neural network model. The learning unit performs learning with respect to the neural network model to which the first quantization has been executed. The bitwidth reducing unit executes a second quantization that applies a second quantization area to a numerical value to be calculated in the neural network model on which learning has been performed in the learning unit. The memory stores the neural network model to which the second quantization has been executed.

Abstract: An apparatus for identification of an input data against one or more learned signals is provided. The apparatus comprising a number of computational cores, each core comprises properties having at least some statistical independency from other of the computational, the properties being set independently of each other core, each core being able to independently produce an output indicating recognition of a previously learned signal, the apparatus being further configured to process the produced outputs from the number of computational cores and determining an identification of the input data based the produced outputs.

Abstract: A system (10) for convolving and adding frames of data comprises a first sensor-display device (14) and a second sensor display device (26). Each sensor display device (14, 26) comprises an array (80) of transmit-receive modules (82). Each transmit-receive module (82) comprises a light sensor element (86), a light transmitter element (84), and a memory bank (90). A radial modulator device (20) is positioned where transmission of light fields comprising frames of data are Fourier transformed. Filters implemented by modulator elements of the radial modulator device (20) convolve the fields of light comprising the frames of data, which are then sensed on a pixel-by-pixel basis by the light sensor elements (86), which accumulate charges, thus sum pixel values of sequential convolved frames of data.

Abstract: Instruction distribution in an array of neural network cores is provided. In various embodiments, a neural inference chip is initialized with core microcode. The chip comprises a plurality of neural cores. The core microcode is executable by the neural cores to execute a tensor operation of a neural network. The core microcode is distributed to the plurality of neural cores via an on-chip network. The core microcode is executed synchronously by the plurality of neural cores to compute a neural network layer.

Abstract: Approaches, techniques, and mechanisms are disclosed for generating, enhancing, applying and updating knowledge neurons for providing decision making information to a wide variety of client applications. Domain keywords for knowledge domains are generated from domain data of selected domain data sources, along with keyword values for the domain keywords, and are used to generate knowledge artifacts for inclusion in knowledge neurons. These knowledge neurons may be enhanced by domain knowledge data sets found in various data sources and used to generate neural responses to neural queries received from the client applications. Neural feedbacks may be used to update and/or generate knowledge neurons.

Abstract: Approaches, techniques, and mechanisms are disclosed for generating, enhancing, applying and updating knowledge neurons for providing decision making information to a wide variety of client applications. Domain keywords for knowledge domains are generated from domain data of selected domain data sources, along with keyword values for the domain keywords, and are used to generate knowledge artifacts for inclusion in knowledge neurons. These knowledge neurons may be enhanced by domain knowledge data sets found in various data sources and used to generate neural responses to neural queries received from the client applications. Neural feedbacks may be used to update and/or generate knowledge neurons.

Abstract: A machine learning method and a machine learning device are provided. The machine learning method includes: receiving an input signal and performing normalization on the input signal; transmitting the normalized input signal to a convolutional layer; and adding a sparse coding layer after the convolutional layer, wherein the sparse coding layer uses dictionary atoms to reconstruct signals on a projection of the normalized input signal passing through the convolutional layer, and the sparse coding layer receives a mini-batch input to refresh the dictionary atoms.

Abstract: A cloud computing system can be configured to generate data models. A model optimizer of the cloud computing system can provision computing resources of the cloud computing system with a data model. A dataset generator of the cloud computing system can generate a synthetic dataset for training the data model. The computing resources can train the data model using the synthetic dataset. The model optimizer can store the data model and metadata of the data model in a model storage. The cloud computing system can receive production data from a data source by a production instance of the cloud computing system using a common file system. The production data can be processed using the data model by the production instance. The computing resources, the dataset generator, and the model optimizer can be hosted by separate virtual computing instances of the cloud computing system.

Abstract: A scalable system and method for completing a model task using a serverless architecture is disclosed. The system may include a model optimizer having one or more memory units for storing instructions and one or more processors. The method may include receiving a request to complete a model task, and retrieving a stored model and a first hyperparameter based on the request. The method may include provisioning first computing resources to a development instance configured to train the retrieved model based on the first hyperparameter and the model task. The method may include receiving, from the development instance, a trained model and a performance metric. The method may include receiving, from a different development instance, a different performance metric associated with a different model, and terminating the development instance based on a determination that the termination condition is satisfied.

Abstract: A model optimizer is disclosed for managing training of models with automatic hyperparameter tuning. The model optimizer can perform a process including multiple steps. The steps can include receiving a model generation request, retrieving from a model storage a stored model and a stored hyperparameter value for the stored model, and provisioning computing resources with the stored model according to the stored hyperparameter value to generate a first trained model. The steps can further include provisioning the computing resources with the stored model according to a new hyperparameter value to generate a second trained model, determining a satisfaction of a termination condition, storing the second trained model and the new hyperparameter value in the model storage, and providing the second trained model in response to the model generation request.

Abstract: A neural network method and apparatus are provided. A processor implemented neural network includes calculating respective individual gradient values for updating a weight of a neural network, calculating a residual gradient value based on an accumulated gradient value obtained by accumulating the individual gradient values and a bit digit representing the weight, tuning the respective individual gradient values to correspond to a bit digit of the residual gradient value, summing the tuned respective individual gradient values, the residual gradient value, and the weight, and updating the weight and the residual gradient value based on a result of the summing to train the neural network.