Abstract: Systems and methods are provided for analyzing, by a computing device, location data associated with a location of the computing device to determine that an image or video captured using a messaging application on the computing device is captured near a food-related venue or event, receiving input related to food associated with the food-related venue or event, sending the image or video and the input related to food associated with the food-related venue or event to a computing system to train a machine learning model for food detection, and updating the messaging application to comprise the trained machine learning model for food detection.

Abstract: A system, method, and computer-readable medium are disclosed for creating image recognition models, which can be operated on smartphone or similar device. The smartphone captures images of hardware in a data center. The captured images are processed to produce a full set of annotated images. The full set is minimized to a simplified set and trained to create a mobile image recognition model implemented by the smartphone or similar device.

Abstract: The present disclosure provides a method and an apparatus for obtaining product training images, and a storage medium. The method includes: obtaining product images on each of product webpages in an e-commerce website, and determining a product feature vector of each product image on the product webpage; dividing the product images on the product webpage, and determining a target image set of the product webpage according to an image dividing result; classifying target image sets of the product webpages according to the average product feature vector to obtain at least one type of image set; and generating the product training images according to the at least one type of image set.

Abstract: Aspects of the detailed technologies concern training and use of neural networks for fine-grained classification of large numbers of items, e.g., as may be encountered in a supermarket. Mitigating false positive errors is an exemplary area of emphasis. Novel network topologies are also detailed—some employing recognition technologies in addition to neural networks. A great number of other features and arrangements are also detailed.

Abstract: A system for generating synthetic training data may include one or processors and a memory in communication with the one or more processors and having a receiving module, a duplication module, and an insertion module. The modules have instructions that when executed by the one or more processors cause the one or more processors to receive original training data being in the form of a three-dimensional point cloud and having one or more original objects formed by at least a portion of the three-dimensional point cloud and annotated with original annotation data, duplicate one of the one or more original objects to generate a synthetic object, and insert the synthetic object within the original training data to generate the synthetic training data.

Abstract: Techniques for detecting label shift and adjusting training data of predictive models in response are provided. In an embodiment, a first machine-learned model is used to generate a predicted label for each of multiple scoring instances. The first machine-learned model is trained using one or more machine learning techniques based on a plurality of training instances, each of which includes an observed label. In response to detecting a shift in observed labels, for each segment of one or more segments in multiple segments, a portion of training data that corresponds to the segment is identified. For each training instance in a subset of the portion of training data, the training instance is adjusted. The adjusted training instance is added to a final set of training data. The machine learning technique(s) are used to train a second machine-learned model based on the final set of training data.

Abstract: Apparatus and methods related to using machine learning to determine depth maps for dual pixel images of objects are provided. A computing device can receive a dual pixel image of at least a foreground object. The dual pixel image can include a plurality of dual pixels. A dual pixel of the plurality of dual pixels can include a left-side pixel and a right-side pixel that both represent light incident on a single dual pixel element used to capture the dual pixel image. The computing device can be used to train a machine learning system to determine a depth map associated with the dual pixel image. The computing device can provide the trained machine learning system.

Abstract: In some embodiments, a method can include capturing images of produce. The method can further include generating simulated images of produce based on the images of produce. The method can further include associating each image of produce from the images of produce and each simulated image of produce from the simulated images of produce with a category indicator, an organic type indicator, and a bag type indicator, to generate a training set. The method can further include training a machine leaning model using the training set such that when the machine learning model is executed, the machine learning model receives an image and generates a predicted category indicator of the image, a predicted organic type indicator of the image, and a predicted bag type indicator of the image.

Abstract: A method and a system for generating an augmented scene graph of an image and for training an explainable knowledge based (KB) visual question answering (VQA) machine learning (ML) model are provided. A scene graph encoding spatial and semantic features of objects and relations between objects in the image is obtained. An augmented scene graph is generated by embedding a knowledge graph to enhance the scene graph. An embedded set of questions and associated answers related to the image are obtained. The KB VQA ML model is trained to provide an answer to a given question related to the image based on the augmented scene graph and the embedded set of questions and associated answers. The KB VQA ML model is trained to retrieve a subgraph linking the question and the associated answer as a potential explanation for the answer.

Abstract: Edge devices utilizing personalized machine learning and methods of operating the same are disclosed. An example edge device includes a model accessor to access a first machine learning model from a cloud service provider. A local data interface is to collect local user data. A model trainer is to train the first machine learning model to create a second machine learning model using the local user data. A local permissions data store is to store permissions indicating constraints on the local user data with respect to sharing outside of the edge device. A permissions enforcer is to apply permissions to the local user data to create a sub-set of the local user data to be shared outside of the edge device. A transmitter is to provide the sub-set of the local user data to a public data repository.

Abstract: Methods, apparatus, systems, and articles of manufacture to simulate sensor data are disclosed. An example apparatus includes a noise characteristic identifier to extract a noise characteristic associated with a feature present in first sensor data obtained by a physical sensor. A feature identifier is to identify a feature present in second sensor data. The second sensor data is generated by an environment simulator simulating a virtual representation of the real sensor. A noise simulator is to synthesize noise-adjusted simulated sensor data based on the feature identified in the second sensor data and the noise characteristic associated with the feature present in the first sensor data.

Abstract: Provided is a process including: writing modelling-object classes using object-oriented modelling of the modelling methods, the modelling-object classes being members of a set of class libraries; writing quality-management classes using object-oriented modelling of quality management, the quality-management classes being members of the set of class libraries; scanning modelling-object classes in the set of class libraries to determine modelling-object class definition information; scanning quality-management classes in the set of class libraries to determine quality-management class definition information; using the modelling-object class definition information and the quality-management class definition information to produce object manipulation functions that allow a quality management system to access methods and attributes of modelling-object classes to manipulate objects of the modelling-object classes; and using the modelling-object class definition information and the quality-management class definitio

Abstract: Embodiments generate a model of demand of a product that includes an optimized feature set. Embodiments receive sales history for the product and receive a set of relevant features for the product and designate a subset of the relevant features as mandatory features. From the sales history, embodiments form a training dataset and a validation dataset and randomly select from the set of relevant features one or more optional features. Embodiments include the selected optional features with the mandatory features to create a feature test set. Embodiments train an algorithm using the training dataset and the feature test set to generate a trained algorithm and calculate an early stopping metric using the trained algorithm and the validation dataset. When the early stopping metric is below a predefined threshold, the feature test set is the optimized feature set.

Abstract: Concepts and technologies disclosed herein are directed to image classification attack mitigation. According to one aspect of the concepts and technologies disclosed herein, a system can obtain an original image and reduce a resolution of the original image to create a reduced resolution image. The system can classify the reduced resolution image and output a first classification. The system also can classify the original image via deep learning image classification and output a second classification. The system can compare the first classification and the second classification. In response to determining that the first classification and the second classification match, the system can output the second classification of the original image. In response to determining that the first classification and the second classification do not match, the system can output the first classification of the original image.

Abstract: Systems and methods which provide postage indicia using mobile communication handsets are shown. According to embodiments, users are enabled to introduce mail pieces into a mail delivery stream which are accepted as having acceptable postage indicia having value associated therewith without the use of traditional metering systems or processor-based postage generation and printing systems. Instead, a mobile communication handset is used to obtain tokens to be used as postage indicia. The tokens may be transmitted to or otherwise printed by appropriate, available printing equipment, such as facsimile machines, network printers, network photo copiers, etc., in response to a request by a mobile communication handset for postage indicia. The foregoing tokens may be printed upon various stock for inclusion with or as a mail piece. The mail piece bearing the token may be placed into the mail stream for processing and delivery using the token as activated as postage indicia.

Abstract: A flexible chip card such as a credit card or a debit card that incorporates a protective layer that precludes unauthorized access to the chip in the chip card. The protective layer is a highly conductive layer that shields the chip and prevents electromagnetic waves that may be emitted by an illicit device from accessing the chip. This protective layer thus prevents any unauthorized persons from obtaining confidential information from the chip card that may then be used to consummate fraudulent transactions or conduct other illicit activities. In one embodiment, the flexible chip card includes one or more stiffening structures to provide rigidity to the chip card.

Abstract: This disclosure relates to technology that provides dynamically configurable access to customized digital content associated with a machine-readable label (“MRL”). A MRL may be designed, printed and distributed to viewers. After distribution, the MRL may be collected by an owner who activates the collectible MRL and associates the MRL with customized content. At a time they are generated, a collectible MRL may be associated with a default scan destination. The default scan destination may be encoded in a data zone of the collectible MRL. The information encoded in a data zone of the collectible MRL may not be changeable after the MRL is distributed or fixed in tangible form. The disclosed technology provides apparatus and methods for customizing content associated with a MRL even after the MRL has been fixed in tangible form.

Abstract: Sheet-like product and method for authenticating a security tag including a section of the sheet-like product. The sheet-like product includes at least one security feature having optical properties that change with the viewing angle and, and at least one marker, wherein each marker is uniquely attributable to a position on the sheet-like product. The position of the at least one security feature on the sheet-like product is predetermined relative to the position of the at least one marker on the sheet-like product.

Abstract: A passive radio frequency identification (RFID) tag includes: a rectifier circuit that rectifies a signal obtained from an antenna and outputs the rectified signal as a DC voltage. A capacitor is connected to an output line of the rectifier circuit. A first regulator circuit generates a first regulator voltage by stabilizing the output DC voltage from the rectifier circuit. A control circuit starts operating when the first regulator voltage is applied, and the control circuit generates a control signal upon receipt of the modulation signal section of the wireless signal. A second regulator circuit generates a second regulator voltage by stabilizing the output DC voltage from the rectifier circuit in response to the control signal and outputs the second regulator voltage to the outside.

Abstract: A bi-directional voltage converter of a smart card includes switching elements connected between an input node and an output node and a start-up transistors whose channel width over channel length is smaller than a channel width over channel length of the switching element. The bi-directional voltage converter stores a driving voltage applied to an output node in a storage capacitor during a booting operation and provides the voltage stored in the storage capacitor to an input node. The bi-directional voltage converter may boost another driving voltage at the input node step-wisely and may perform bi-directional voltage converting with reduced occupied area and high efficiency.

Abstract: A tracking tag system includes a tag case, a first tag case back plate and a second tag case back plate. The tag case has a mating surface and is configured to house a PCB having a transmitter and circuitry for controlling the transmitter. The first tag case back plate is configured to mate with the tag case to enclose and totally protect the PCB, the transmitter, the circuitry and a battery of a first type against ingress of water during immersion and against dust ingress. The second tag case back plate configured to enclose and totally protect the PCB, the transmitter, the circuitry and a battery of a second type against ingress of water during immersion and against dust ingress.

Abstract: According to one embodiment, a reading apparatus includes a shielding body, an antenna, and a reader and writer. The shielding body is formed in a box shape with an upper opening, to place an accommodating body and to shield radio waves. The antenna is provided in the shielding body to receive information from an RFID tag attached to a product that is passing through the opening. The reader and writer is connected to the antenna to read information of the product from the information received by the antenna.

Abstract: Apparatus and methods disclosed herein provide technical solution for on-demand manufacturing of a payment instrument that includes an integrated circuit chip. Apparatus and methods provide technical solutions for securely validating and activating the manufactured payment instrument. The customer may submit a request to manufacture a payment instrument through an automated teller machine (“ATM”), online banking channel or mobile banking channel. Apparatus and methods allow a customer to manufacture a payment instrument at home using a 3D printer or use a 3D printer installed at an ATM. Using a secure validation methods, the newly manufactured payment instrument may be activated at an ATM.

Abstract: The present invention relates to a prelaminate for an electronic card, wherein at least a first group of pads is formed from a metal plate formed from a piece comprising a central part and branches extending from the central part, the branches of the metal plate forming the pads of the first group. The invention also relates to a method for producing such a prelaminate and an electronic card comprising such a prelaminate.

Abstract: An identification element (50) for a shank tool has an elastic inner ring (10) and an outer ring (20) arranged around the inner ring (10), said outer ring having a recess formed annularly in the outer ring. An RFID transponder is arranged in the recess. The inner ring (10) of the identification element (50) encloses a shank of the shank tool.

Abstract: A biological sample storage container and a dual chip wireless identification tag thereof are provided. The dual chip wireless identification tag includes a substrate, an antenna structure, a first chip, and a second chip. The antenna structure is disposed on the substrate, and includes two radiation parts and two matching parts. The two matching parts are connected between the two radiation parts, the first chip is coupled to one of the matching parts, and the second chip is coupled to the other one of the matching parts.

Abstract: The present disclosure relates to an automotive virtual personal assistant configured to provide intelligent support to a user, mindful of the user environment both in and out of a vehicle. Further, the automotive virtual personal assistant is configured to contextualize user-specific vehicle-based and cloud-based data to intimately interact with the user and predict future user actions. Vehicle-based data may include spoken natural language, visible and infrared camera video, as well as on-board sensors of the type commonly found in vehicles. Cloud-based data may include web searchable content and connectivity to personal user accounts, fully integrated to provide an attentive and predictive user experience. In contextualizing and communicating these data, the automotive virtual personal assistant provides improved safety and an enhanced user experience.

Abstract: A method for recommending an action to a user of a user device includes receiving first user action data corresponding to a first user action and receiving second user action data corresponding to a second user action. The method also includes generating, based on the first user action data and the second user action data and using a feedforward artificial neural network, a recommendation for a next user action. The method also includes causing the recommendation for the next user action to be communicated to the user device.

Abstract: The specification discloses a question answer matching method, system and computer storage medium. The method comprises: transforming the user query and one of one or more suggested answers corresponding to the user query by using a pre-trained word vector to obtain vector representations of the user query and the one of one or more suggested answers corresponding to the user query; performing a convolutional operation on the vector representations of the user query and the one of one or more suggested answers, respectively, to extract features; and mapping convolution results of the vector representations of the user query and the vector expression of the one of one or more suggested answers into a sample annotating space, to obtain a matching result of the user query.

Abstract: A state machine engine having a program buffer. The program buffer is configured to receive configuration data via a bus interface for configuring a state machine lattice. The state machine engine also includes a repair map buffer configured to provide repair map data to an external device via the bus interface. The state machine lattice includes multiple programmable elements. Each programmable element includes multiple memory cells configured to analyze data and to output a result of the analysis.

Abstract: Recurrent neural networks, and methods therefor, are provided with diagonal and programming fluctuation to find energy global minima. The method may include storing the matrix of weights in memory cells of a crossbar array of a recursive neural network prior to operation of the recursive neural network; altering the weights according to a probability distribution; setting the weights to non-zero values in at least one of the memory cells in a diagonal of the memory cells in the crossbar array; and operating the recursive neural network.

Abstract: Guided character string alteration can be performed by obtaining an original character string and a plurality of altered character strings, traversing the original character string with a first Long Short Term Memory (LSTM) network to generate, for each character of the original character string, a hidden state of a partial original character string up to that character, and applying, during the traversing, an alteration learning process to each hidden state of a partial original character string to produce an alteration function for relating partial original character strings to partial altered character strings.

Abstract: Examples described herein utilize multi-layer neural networks to decode encoded data (e.g., data encoded using one or more encoding techniques). The neural networks have nonlinear mapping and distributed processing capabilities which are advantageous in many systems employing the neural network decoders. In this manner, neural networks described herein are used to implement error code correction (ECC) decoders.

Abstract: Techniques are provided for training machine learning model. According to one aspect, a training data is received by one or more processing units. The machine learning model is trained based on the training data, wherein the training comprises: optimizing the machine learning model based on stochastic gradient descent (SGD) by adding a dynamic noise to a gradient of a model parameter of the machine learning model calculated by the SGD.

Abstract: Graphical elements in a user interface (UI) may be detected in robotic process automation (RPA) using convolutional neural networks (CNNs). Such processes may be particularly well-suited for detecting graphical elements that are too small to be detected using conventional techniques. The accuracy of detecting graphical elements (e.g., control objects) may be enhanced by providing neural network-based processing that is robust to changes in various UI factors, such as different resolutions, different operating system (OS) scaling factors, different dots-per-inch (DPI) settings, and changes due to UI customization of applications and websites, for example.

Abstract: The ability to rapidly identify symmetry and anti-symmetry is an essential attribute of intelligence. Symmetry perception is a central process in human vision and may be key to human 3D visualization. While previous work in understanding neuron symmetry perception has concentrated on the neuron as an integrator, the invention provides the coincidence detecting property of the spiking neuron can be used to reveal symmetry density in spatial data. A synchronized symmetry-identifying spiking artificial neural network enables layering and feedback in the network. The network of the invention can identify symmetry density between sets of data and present a digital logic implementation demonstrating an 8×8 leaky-integrate-and-fire symmetry detector in a field-programmable gate array. The efficiency of spiking neural networks can be harnessed to rapidly identify symmetry in spatial data with applications in image processing, 3D computer vision, and robotics.

Abstract: In an example, an apparatus comprises a plurality of execution units comprising and logic, at least partially including hardware logic, to traverse a solution space, score a plurality of solutions to a scheduling deep learning network execution, and select a preferred solution from the plurality of solutions to implement the deep learning network. Other embodiments are also disclosed and claimed.

Abstract: Disclosed are an artificial neural network device and a method of operating the same. The artificial neural network device includes an operation part performing an artificial neural network operation on an input feature map and a classification part performing a classifying operation on the input feature map based on the artificial neural network operation of the operation part. The operation part includes an XNOR operation circuit performing an XNOR operation on the input feature map and a filter and a binarizing circuit performing a binarization operation based on the result of the XNOR operation of the XNOR operation circuit. Accordingly, the artificial neural network device is miniaturized and performs the operation at high speed.

Abstract: Disclosed is neural network circuitry having a first plurality of logic cells that is interconnected to form neural network computation units that are configured to perform approximate computations. The neural network circuitry further includes a second plurality of logic cells that is interconnected to form a controller hierarchy that is interfaced with the neural network computation units to control pipelining of the approximate computations performed by the neural network computational units. In some embodiments the neural network computation units include approximate multipliers that are configured to perform approximate multiplications that comprise the approximate computations. The approximate multipliers include preprocessing units that reduce latency while maintaining accuracy.

Abstract: A neural processor circuit including one or more planar engine circuits that perform non-convolution operations in parallel with convolution operations performed by one or more neural engine circuits. The neural engine circuits perform the convolution operations on neural input data corresponding to one or more neural engine tasks to generate neural output data. The planar engine circuits perform non-convolution operations on planar input data corresponding to one or more planar engine tasks to generate planar output data. A data processor circuit in the neural processor circuit addresses data dependency between the one or more neural engine tasks and the one or more planar engine tasks by controlling reading of the neural output data as the planar input data by the planar engine circuits or reading of the planar output data as the neural input data by the neural engine circuits.

Abstract: A memristive device is described. The memristive device includes a first layer having a first plurality of conductive lines, a second layer having a second plurality of conductive lines, and memristive interlayer connectors. The first and second layers differ. The first and second pluralities of conductive lines are each lithographically defined. The first and second pluralities of conductive lines are insulated from each other. The memristive interlayer connectors are memristively coupled with a first portion of the first plurality of conductive lines and memristively coupled with a second portion of the second plurality of conductive lines. The memristive interlayer connectors are thus sparsely coupled with the first and second pluralities of conductive lines. Each memristive interlayer connector includes a conductive portion and a memristive portion.

Abstract: An analog computing method includes the steps of: (a) generating a biasing current (IWi) using a constant gm bias circuit operating in the subthreshold region for ultra-low power consumption, wherein gm is generated by PMOS or NMOS transistors, the circuit including a switched capacitor resistor; and (b) multiplying the biasing current by an input voltage using a differential amplifier multiplication circuit to generate an analog voltage output (VOi). In one or more embodiments, the method is used in a vision application, where the biasing current represents a weight in a convolution filter and the input voltage represents a pixel voltage of an acquired image.

Abstract: A function creation method is disclosed. The method comprises defining one or more database function inputs, defining cluster processing information, defining a deep learning model, and defining one or more database function outputs. A database function is created based at least in part on the one or more database function inputs, the cluster set-up information, the deep learning model, and the one or more database function outputs. In some embodiments, the database function enables a non-technical user to utilize deep learning models.

Abstract: A signal processing apparatus includes one or more processors. The one or more processors perform first-type signal processing on an input signal using a neural network, and output a first-type output signal. The one or more processors convert the first-type output signal into a second-type output signal for calculating a first-type loss related to accuracy of second-type signal processing performed by another signal processing device The one or more processors calculate the first-type loss based on the second-type output signal and a correct signal. The one or more processors optimize parameters of the neural network based on the first-type loss.

Abstract: A Static Random Access Memory (SRAM) device in a binary neural network is provided. The SRAM device includes an SRAM inference engine having an SRAM computation architecture with a forward path that include multiple SRAM cells. The multiple SRAM cells are configured to form a chain of SRAM cells such that an output of a given one of the multiple SRAM cells is an input to a following one of the multiple SRAM cells. The SRAM computation architecture is configured to compute a prediction from an input.

Abstract: Systems and methods are disclosed for applying neural networks in resource-constrained environments. A system may include a sensor located in a resource-constrained environment configured to generate sensor data of the resource-constrained environment. The system may also include a first computing device not located in the resource-constrained environment configured to produce a neural network structure based on the sensor data. The system may further include a second computing device located in the resource-constrained environment configured to provide the sensor data as input to the neural network structure. The second computing device may be further configured to determine a state of the resource-constrained environment based on the input of the sensor data to the neural network structure.

Abstract: A hardware-implemented multi-layer perceptron model calculation unit includes: a processor core to calculate output quantities of a neuron layer based on input quantities of an input vector; a memory that has, for each neuron layer, a respective configuration segment for storing configuration parameters and a respective data storage segment for storing the input quantities of the input vector and the one or more output quantities; and a DMA unit to successively instruct the processor core to: calculate respective neuron layers based on the configuration parameters of each configuration segment, calculate input quantities of the input vector defined thereby, and store respectively resulting output quantities in a data storage segment defined by the corresponding configuration parameters, the configuration parameters of configuration segments successively taken into account indicating a data storage region for the resulting output quantities corresponding to the data storage region for the input quantities for

Abstract: A parameter training method for a convolutional neural network (CNN) for classifying image type data representative of a biometric trait. The method includes the implementation, by a data processor of a server, the steps of (a) for at least one data item from an already classified training database, generation of several alternate versions of this data each by application of at least one transformation chosen from a set of reference transformations satisfying a statistical distribution of transformations observed in the training database and (b) training the parameters of the CNN, from the already classified training database augmented with said alternate versions of the data.

Abstract: The present invention discloses a hierarchical highly heterogeneous distributed system based deep learning application optimization framework and relates to the field of deep learning in the direction of computational science. The hierarchical highly heterogeneous distributed system based deep learning application optimization framework comprises a running preparation stage and a running stage. The running preparation stage is used for performing deep neural network training. The running stage performs task assignment to all kinds of devices in the distributed system and uses a data encryption module to perform privacy protection to user sensitive data.

Abstract: Embodiments of the subject technology for deep learning based reservoir modelling provides for receiving input data comprising information associated with one or more well logs in a region of interest. The subject technology determines, based at least in part on the input data, an input feature associated with a first deep neural network (DNN) for predicting a value of a property at a location within the region of interest. Further, the subject technology trains, using the input data and based at least in part on the input feature, the first DNN. The subject technology predicts, using the first DNN, the value of the property at the location in the region of interest. The subject technology utilizes a second DNN that classifies facies based on the predicted property in the region of interest.