Patents Issued in February 1, 2024
-
Publication number: 20240037348Abstract: A method of transforming data including receiving data in a first language specific form, converting the data in the first language specific form to a language agnostic form, storing the data in the language agnostic form, converting the data in the language agnostic form to at least one second language specific form and exporting, on demand, the data in at least one of the at least one second language specific form.Type: ApplicationFiled: October 12, 2023Publication date: February 1, 2024Applicant: LISUTO KKInventors: Nir PLATEK, Pavel ZASLAVSKY
-
Publication number: 20240037349Abstract: Provided are a model training method and apparatus, a machine translation method and apparatus, a device, and a storage medium. The model training method includes the steps described below. Through a neural network pruning technique, a respective influence degree of each parameter in multiple parameters in a first translation model on a translation result in a first field is determined to obtain at least one first parameter and at least one second parameter. By using the first corpus of the first field, the at least one first parameter is trained obtain the second translation model, and the at least one second parameter remains unchanged. Similarity between a translation result of the second translation model in the first field and a translation result of the first translation model in the first field meets a preset condition.Type: ApplicationFiled: November 17, 2021Publication date: February 1, 2024Inventors: Chengqi ZHAO, Jianze LIANG, Mingxuan WANG, Lei LI
-
Publication number: 20240037350Abstract: Some implementations disclosed herein provide techniques and arrangements to enable translating language characters in media content. For example, some implementations receive a user selection of a first portion of media content. Some implementations disclosed herein may, based on the first portion, identify a second portion of the media content. The second portion of the media content may include one or more first characters of a first language. Some implementations disclosed herein may create an image that includes the second portion of the media content and may send the image to a server. Some implementations disclosed herein may receive one or more second characters of a second language corresponding to a translation of the one or more first characters of the first language from the server.Type: ApplicationFiled: October 13, 2023Publication date: February 1, 2024Applicant: Microsoft Technology Licensing, LLCInventors: Jun DU, Lei SUN, Jian SUN, Qiang HUO
-
Publication number: 20240037351Abstract: An example communication device is disclosed herein. The communication device includes a magnetic-field interface; a near-field radio frequency (RF) interface; a far-field RF interface; and a controller. The controller is configured to place the communication device in a deep sleep mode, and in response to receiving a wake-up signal at the near-field RF interface, transition the communication device from the deep sleep mode to an awake mode for a period of time. If an activation signal is received during the period of time, the controller can transition the communication device from the awake mode to a fully functional mode, and if the activation signal is not received during the period of time, the controller can transition the communication device from the awake mode to the deep sleep mode.Type: ApplicationFiled: October 13, 2023Publication date: February 1, 2024Inventors: Alexander Mueggenborg, Edward A. Richley
-
Publication number: 20240037352Abstract: A method for security tag removal includes providing a terminal comprising a receptacle, which is configured to receive a magnetic security tag that is attached to an item of merchandise. A wireless transceiver in the terminal reads identification data that is encoded in the security tag while the security tag is in the receptacle. A query is transmitted from the terminal to a server with respect to the identification data. In response to the query, an authorization is received from the server with respect to the item of merchandise to which the security tag is attached. In response to the authorization, a magnet in the terminal is actuated so as to release the security tag from the item of merchandise.Type: ApplicationFiled: October 16, 2023Publication date: February 1, 2024Inventors: Ori David Levi, Barak Amtanani
-
Publication number: 20240037353Abstract: A space efficient automated processing system for processing objects is disclosed. The processing system includes an input conveyance system for moving objects from an input area in at least an input conveyance vector that includes an input conveyance horizontal direction component and an input conveyance vertical direction component, a perception system for receiving objects from the input conveyance system and for providing perception data regarding an object, a primary transport system for receiving the object from the perception system and for providing transport of the object along at least a primary transport vector including an primary transport horizontal component and a primary transport vertical component that is generally opposite the input conveyance horizontal direction component, and at least two secondary transport systems, each of which receives the object from the primary transport system and moves the object in either of reciprocal directions.Type: ApplicationFiled: October 5, 2023Publication date: February 1, 2024Inventors: Thomas WAGNER, Kevin Ahearn, John Richard Amend, JR., Benjamin Cohen, Michael Dawson-Haggerty, William Hartman Fort, Christopher Geyer, Victoria Hinchey, Jennifer Eileen King, Thomas Koletschka, Michael Cap Koval, Kyle Maroney, Matthew T. Mason, William Chu-Hyon Mcmahan, Gene Temple Price, Joseph Romano, Daniel Smith, Siddhartha Srinivasa, Prasanna Velagapudi, Thomas Allen
-
Publication number: 20240037354Abstract: Handheld barcode readers and assemblies are disclosed herein. An example handheld barcode reader includes a housing defining a head portion and a base portion, a vision camera positioned in the base portion, a barcode reading module positioned at least partially in the head portion, and a controller in communication with the barcode reading module and the vision camera. The vision camera has a first FOV directed through a base window in the base portion and the barcode reading module has a second FOV directed through a scan window in the head portion. The controller is configured to decode barcodes read by the barcode reading module, receive captured images from the vision camera, and synchronize the barcode reading module and the vision camera such that the vision camera does not capture images when the barcode reading module is active.Type: ApplicationFiled: October 16, 2023Publication date: February 1, 2024Inventors: Darran Michael Handshaw, Edward Barkan, Ronald Steven Ethe
-
Publication number: 20240037355Abstract: A system for optically reading an encoded pattern on a patterned test strip or other patterned target placed in proximity and moved with respect to an optical code reader. The reader can include an LED light source. A transmit baffle located between the LED light source and the target can define a transmit aperture for passing light originating from the LED light source to a target location on the patterned target. A photodetector arranged to receive light associated with the target location can produce a response signal providing information about the encoded pattern at the target location. A receive baffle located between the photodetector and the target can define a receive aperture for passing light from the target to the photodetector. Signal processing circuitry can be coupled to the photodetector to receive the response signal, and to decode information about the encoded pattern.Type: ApplicationFiled: December 17, 2021Publication date: February 1, 2024Inventor: Shrenik Deliwala
-
Publication number: 20240037356Abstract: A reading device includes: a handy scanner; an adjuster; and a pedestal. The handy scanner includes a head incorporating a reading sensor and a handle which is a gripping part having one end portion connected to the head, and configured to read a code symbol image by the reading sensor. The adjuster is attached to the head and surrounds a readable angular area of the reading sensor. The pedestal is a portion to which the adjuster is detachably coupled, and holds the adjuster in such a direction that the readable angular area of the reading sensor is substantially equal in an up-down direction with respect to a horizontal direction.Type: ApplicationFiled: May 26, 2023Publication date: February 1, 2024Inventor: Kento Kawata
-
Publication number: 20240037357Abstract: An example barcode reader includes a housing having a window. An imaging assembly is recessed within the housing, has an imaging field-of-view (FOV) directed toward the window, and includes an imager configured to capture an image of an object positioned within a product scanning region. An illumination assembly is recessed within the housing and has an illumination FOV directed toward the window and overlaps the imaging FOV by at least 70% from zero to 3 inches from the window. A controller is configured to receive a captured image from the imaging assembly and to process the captured image to: determine if the object is a product for sale or a hand of a user; decode a barcode if the object is a product for sale; and identify elements of the hand of the user for identification if the object is the hand of the user.Type: ApplicationFiled: July 29, 2022Publication date: February 1, 2024Inventors: Darran Michael Handshaw, Stefanie Handshaw
-
Publication number: 20240037358Abstract: At least some embodiments of the present invention are directed to barcode readers having a housing with upper and lower portions, and a weigh platter. Additionally, the barcode readers include a first imaging assembly having a first imaging sensor, the first imaging assembly having a first field of view (FOV) directed through at least one of the substantially horizontal window or the substantially upright window of the housing, and a second imaging assembly having a second imaging sensor, the second imaging assembly having a second FOV and being positioned near the distal edge of the weigh platter that is opposite the upper portion of the barcode reader housing.Type: ApplicationFiled: July 29, 2022Publication date: February 1, 2024Inventors: Edward Barkan, Darran Michael Handshaw, Mark Drzymala
-
Publication number: 20240037359Abstract: Scanning systems are disclosed herein. An example scanning system includes a cradle and a barcode scanner. The cradle includes a cradle cavity portion, a cradle controller, a force sensor communicatively coupled therewith, and a first securing feature positioned at or near the cradle cavity portion and being communicatively coupled with the cradle controller. The barcode scanner includes a housing having a scanner housing cavity, an imaging assembly adapted to capture an image of an environment appearing in a field of view (FOV) and being at least partially disposed within the scanner housing cavity, a scanner controller adapted to control operation of the imaging assembly, and a second securing feature. In response to the force sensor sensing vibration exceeding a threshold value and/or the cradle being mounted in a predetermined orientation, the first and second securing features selectively interact to retain the barcode scanner within the cradle cavity portion.Type: ApplicationFiled: July 28, 2022Publication date: February 1, 2024Inventor: Christopher P. Klicpera
-
Publication number: 20240037360Abstract: An image forming apparatus comprises a communicator that receives a print command of image data from a user terminal via a network; an image former that forms an image on the basis of the image data; and one or more controllers that control the communicator and the image former, wherein when a time-out period set in advance has elapsed without receiving a transmission completion notification of a print command from the user terminal after receiving the print command from the user terminal, the one or more controllers determine whether image data related to the print command is complete or not in accordance with a determination procedure set in advance, and when the one or more controllers determine that the image data is complete, the one or more controllers cause the image former to form an image based on the image data, while when the one or more controllers determine that the image data is not complete, the one or more controllers withhold execution of the print command.Type: ApplicationFiled: July 21, 2023Publication date: February 1, 2024Applicant: SHARP KABUSHIKI KAISHAInventor: Katsuya Kitayama
-
Publication number: 20240037361Abstract: A bitmap processing system for creating digital documents on a digital printing press by a fulfiller includes a fulfiller operated raster image processor that receives customer application Page Description Language (PDL) job files containing information for creating the digital documents and generates bitmaps in accordance with the files and a fulfiller operated editor that modifies the bitmaps so that the resulting documents process more efficiently in production steps downstream of a printing process.Type: ApplicationFiled: October 9, 2023Publication date: February 1, 2024Inventors: Frank Delfer, Marc J. Fagan, Charles Preston
-
Publication number: 20240037362Abstract: A method of generating a barcode pattern includes: identifying a silhouette comprising a boundary within which a barcode pattern is to be printed; generating a barcode pattern comprising multiple instances of a two-dimensional barcode, in which the instances include barcodes of at least two different sizes; and generating a print file that includes information for printing the barcode pattern within the silhouette.Type: ApplicationFiled: July 24, 2023Publication date: February 1, 2024Inventor: Schuyler Justin Gordon Van Sickle
-
Publication number: 20240037363Abstract: A method for manufacturing a card body for a chip card, includes the steps of: supplying a metallic base body with two opposite main faces and a circumferential peripheral face connecting the two main faces, wherein in the base body a module opening for receiving a chip module has already been produced or will still be produced in a module opening zone, and producing a slot on the peripheral face between the two main faces. The slot is formed from the peripheral face up to the module opening or up to the module opening zone. An entry angle (?) of the slot into at least one of the two main faces is not equal to ninety degrees with respect to the main face.Type: ApplicationFiled: December 15, 2021Publication date: February 1, 2024Inventor: Michael BALDISCHWEILER
-
Publication number: 20240037364Abstract: An asset tag adapted to be mounted to an asset. The asset tag comprises a first component encoded with a first ID unique to the asset tag, the first component having a first wireless interface and transmitting first broadcast signals via said first wireless interface over a first range, the first broadcast signals including the first ID. The asset tag further comprises a user-actuatable button. The asset tag further comprises processing circuitry, coupled to the button and to at least the first wireless interface. The processing circuitry is configured for (i) determining whether a predetermined gesture has been performed by a user using the button and (ii) if the predetermined gesture has been performed, transmitting via the first wireless interface to a wireless access point a restock message, the restock message including the first ID and indicating that restocking is required of assets corresponding to the first ID.Type: ApplicationFiled: October 16, 2023Publication date: February 1, 2024Inventors: Dean Charles HENRY, John VENTER
-
Publication number: 20240037365Abstract: Mounts for securing a device to a flexible object comprise a flexible substrate configured to be operatively coupled to the flexible object and a housing coupled to the flexible substrate and having an internal void that is sized to receive the device.Type: ApplicationFiled: May 11, 2023Publication date: February 1, 2024Inventors: Casey Hopkins, Jacob Hull
-
Publication number: 20240037366Abstract: An RFID mesh label configured to be integrally incorporated within a vulcanized tire and to further provide unique identifier(s) and/or other information about the vulcanized tire during and/or post-vulcanization, the RFID mesh label including a face layer configured to be positioned adjacent or flush to an outer surface of the vulcanized tire; an RFID layer positioned underneath the face layer, the RFID layer having an RFID device that is configured to provide unique identifier(s) and/or other information about the vulcanized tire upon being read with an RFID reader; and a mesh backing overlying the RFID layer and adapted to be integrally incorporated in a vulcanized tire after subjecting a green tire to a vulcanization process.Type: ApplicationFiled: October 16, 2023Publication date: February 1, 2024Inventors: Jos Uijlenbroek, Michael E. Borgna
-
Publication number: 20240037367Abstract: Apparatuses, systems, and techniques to infer a sequence of actions to perform using one or more neural networks trained, at least in part, by optimizing a probability distribution function using a cost function, wherein the probability distribution represents different sequences of actions that can be performed. In at least one embodiment, a model predictive control problem is formulated as a Bayesian inference task to infer a set of solutions.Type: ApplicationFiled: April 12, 2023Publication date: February 1, 2024Inventors: Alexander Conrad Lambert, Adam Harper Fishman, Dieter Fox, Byron Boots, Fabio Tozeto Ramos
-
Publication number: 20240037368Abstract: Methods, systems, and apparatus, including computer programs encoded on computer-storage media, for managing display devices using machine learning. In some implementations, a system receives image data representing an image provided for presentation by a display device. The system processes the image data using a machine learning model that has been trained to evaluate status of display devices based on input of image data corresponding to the display devices. The system selects a classification for a status of the display device based on the output that the machine learning model generated based on the image data. The system provides an output indicating the selected classification over the communication network in response to receiving the image data.Type: ApplicationFiled: July 28, 2022Publication date: February 1, 2024Inventors: Amit Arora, Sonia Thakur, Namrata Walanj
-
Publication number: 20240037369Abstract: Disclosed are a method and an apparatus for multiple-beat detection using electrocardiogram global feature vectors. This method and apparatus extracts global features of each electrocardiogram wave, and extracts and learns, using the extracted global features as input vectors, a pattern of global features of a consecutive electrocardiogram wave by applying an attention mechanism to a weighted feature matrix in consideration of the degree of contribution of each feature to detect multiple beats.Type: ApplicationFiled: December 23, 2021Publication date: February 1, 2024Applicant: SEERSTECHNOLOGY CO.,LTD.Inventors: Youngshin LEE, Heeseok SONG, Yunkwan KIM
-
Publication number: 20240037370Abstract: A system and method are disclosed herein. The system includes one or more processors and a memory having programming instructions stored thereon, which, when executed by the one or more processors, performs operations. The operations include retrieving historical account activity. The operations further include constructing a training data set that includes the historical inflow data, the historical outflow data, and known forecast information from the historical account activity. The operations further include generating a combined prediction model configured to forecast future inflow activity and future outflow activity. The operations further include receiving current inflow activity, current outflow activity, and current balance information for a user.Type: ApplicationFiled: October 11, 2023Publication date: February 1, 2024Inventors: Jason Chun Sing Chan, Richard Saunders, Ryan Murray, Yuchen Hua, Adrian Yu Hin Lam, Sri Chaitanya Somanchi, Noam Katz
-
Publication number: 20240037371Abstract: One example method includes receiving, by a machine learning (“ML”) model of a conference client application, audio signals received from a microphone of a client device, the client device connected to a virtual meeting via the conference client application, the virtual meeting hosted by a virtual conference provider; determining, by the ML model, a plurality of candidate reactions associated with the audio signals, the ML comprising a plurality of convolutional neural network (“CNN”) layers and at least one fully connected layer; selecting a reaction from the plurality of candidate reactions; and transmitting the reaction to the virtual conference provider.Type: ApplicationFiled: July 26, 2022Publication date: February 1, 2024Applicant: Zoom Video Communications, Inc.Inventors: Yuhui Chen, Qiang Gao, Zhaofeng Jia, Rongrong Liu
-
Publication number: 20240037372Abstract: The present invention relates to machine learning (ML) explainability (MLX). Herein are techniques for a novel relevance propagation rule in layer-wise relevance propagation (LRP) for feature attribution-based explanation (ABX) for a reconstructive autoencoder. In an embodiment, a reconstruction layer of a reconstructive neural network in a computer generates a reconstructed tuple that is based on an original tuple that contains many features. A reconstruction residual cost function calculates a reconstruction error that measures a difference between the original tuple and the reconstructed tuple. Applied to the reconstruction error is a novel reconstruction relevance propagation rule that assigns a respective reconstruction relevance to each reconstruction neuron in the reconstruction layer. Based on the reconstruction relevance of the reconstruction neurons, a respective feature relevance of each feature is determined, from which an ABX explanation may be automatically generated.Type: ApplicationFiled: July 26, 2022Publication date: February 1, 2024Inventors: Kenyu Kobayashi, Arno Schneuwly, Renata Khasanova, Matteo Casserini, Felix Felix Schmidt
-
Publication number: 20240037373Abstract: Aspects of the disclosure are directed to jointly searching machine learning model architectures and hardware architectures in a combined space of models, hardware, and mapping strategies. A search strategy is utilized where all models, hardware, and mappings are evaluated together at once via weight sharing and a supernetwork. A multi-objective reward function is utilized with objectives for quality, performance, power, and area.Type: ApplicationFiled: July 28, 2022Publication date: February 1, 2024Inventors: Sheng Li, Norman Paul Jouppi, Garrett Axel Andersen, Quoc V. Le, Liqun Cheng, Parthasarathy Ranganathan
-
Publication number: 20240037374Abstract: Systems, computer program products, and computer-implemented methods for determining relationships between one or more outputs of a first model and one or more inputs of a second model that collectively represent a real world system, and chaining the models together. For example, the system described herein may determine how to chain a plurality of models by training an artificial intelligence system using the nodes of the models such that the trained artificial intelligence system predicts related output and input node connections. The system may then link related nodes to chain the models together. The systems, computer program products, and computer-implemented methods may thus, according to various embodiments, enable a plurality of discrete models to be optimally chained.Type: ApplicationFiled: October 9, 2023Publication date: February 1, 2024Inventors: Jesse Rickard, Andrew Floren, Timothy Slatcher, David Skiff, Thomas McArdle, David Fowler, Aravind Baratha Raj
-
Publication number: 20240037375Abstract: An artificial intelligence (AI)-based knowledge distillation and paper production computing system processes instructions to use machine learning models to automatically review papers from a large corpus of papers and distill knowledge using science of science methods and AI-based modeling techniques. The AI-based knowledge distillation and paper production computing system processes instructions to leverage network science and machine learning tools to analyze papers with respect to a given topic to find relevant scientific publications, organize and group publications based on topic similarity and relation to the topic in general, and distill and summarize the message and content of these publications into a coherent set of statements.Type: ApplicationFiled: December 17, 2021Publication date: February 1, 2024Inventors: Dashun Wang, Nima Dehmamy, Lu Liu, Woo Seong Jo
-
Publication number: 20240037376Abstract: A signal processing apparatus executes a convolution operation of predetermined layers constituting a neural network; and transfers first form data to be stored in a storage. The apparatus executes, on output data outputted from a convolution operation of a first layer among the predetermined layers, an arithmetic operation of a compression layer that is configured by a neural network and compresses data, and outputs the first form data to be transmitted to the storage. The apparatus further executes, on the first form data stored in the storage, an arithmetic operation of a restoration layer that is configured by a neural network and restores pre-compression data, and outputs input data to be inputted to a convolution operation of a second layer among the predetermined layers.Type: ApplicationFiled: July 18, 2023Publication date: February 1, 2024Inventors: Hayato OURA, Takayuki KOMATSU, Takaaki YOKOI
-
Publication number: 20240037377Abstract: A method and apparatus are provided. The method includes reordering a plurality of filters, then based on a result of the reordering, compressing weights, among a plurality of weights of the plurality of filters, resulting in some of the plurality of weights being uncompressed weights, generating a plurality of operation unit maps by mapping the uncompressed weights to respective operation units according to a predetermined bulk unit, and mapping the plurality of operation unit maps to an array.Type: ApplicationFiled: June 30, 2023Publication date: February 1, 2024Applicants: Samsung Electronics Co., Ltd., Seoul National University R&DB FoundationInventors: Hoon SHIN, Jae Wook LEE, Rihae PARK, Yeonhong PARK, Seung Yul LEE, Hyunseung LEE
-
Publication number: 20240037378Abstract: Systems, apparatuses and methods may provide for technology that identifies an embedding table associated with a neural network. The neural network is associated with a plurality of compute nodes. The technology further identifies a number of entries of the embedding table, and determines whether to process gradients associated with the embedding table as dense gradients or sparse gradients based on the number of entries.Type: ApplicationFiled: December 24, 2020Publication date: February 1, 2024Applicant: Intel CorporationInventors: Guokai Ma, Jiong Gong, Dhiraj Kalamkar, Rachitha Prem Seelin, Hongzhen Liu, Akshay Jain, Liangang Zhang
-
Publication number: 20240037379Abstract: A server system with AI accelerator apparatuses using in-memory compute chiplet devices. The system includes a plurality of multiprocessors each having at least a first server central processing unit (CPU) and a second server CPU, both of which are coupled to a plurality of switch devices. Each switch device is coupled to a plurality of AI accelerator apparatuses. The apparatus includes one or more chiplets, each of which includes a plurality of tiles. Each tile includes a plurality of slices, a CPU, and a hardware dispatch device. Each slice can include a digital in-memory compute (DIMC) device configured to perform high throughput computations. In particular, the DIMC device can be configured to accelerate the computations of attention functions for transformer-based models (a.k.a. transformers) applied to machine learning applications. A single input multiple data (SIMD) device configured to further process the DIMC output and compute softmax functions for the attention functions.Type: ApplicationFiled: October 13, 2023Publication date: February 1, 2024Inventors: Jayaprakash BALACHANDRAN, Akhil ARUNKUMAR, Aayush ANKIT, Nithesh Kurella, Sudeep Bhoja
-
Publication number: 20240037380Abstract: An analog neuromphric circuit is disclosed having an input layer, a liquid layer, and an output layer each with resistive memory crossbar configurations to identify a temporal signal for a duration of time. The input layer encodes input layer spiking neurons based on encoding signals generated from input voltages applied an input layer resistive memory crossbar configuration. The liquid layer counts each spike generated by liquid layer spiking neurons for the duration of time based on liquid layer signals generated from the input spiking neuron voltages generated from each input layer spiking neurons applied to a liquid layer resistive memory crossbar configuration. The output layer identifies the temporal signal for the duration of time based on output voltages generated from the counting voltages generated from each count of each spike generated by the liquid layer spiking neurons for the duration of time applied to an output resistive memory crossbar configuration.Type: ApplicationFiled: July 26, 2023Publication date: February 1, 2024Inventors: Alex Henderson, Chris Yakopcic, Tarek M. Taha
-
Publication number: 20240037381Abstract: According to one aspect of the present invention, a ternary neural network accelerator device includes a first semiconductor device comprising a first source terminal, a first drain terminal, and a first gate terminal, a second semiconductor device comprising a second source terminal, a second drain terminal, and a second gate terminal, a first searching line connected to the first drain terminal, a second searching line connected to the second drain terminal, and a matching line commonly connected to the first source terminal and the second source terminal, wherein ternary weight and ternary input are each set by either of a first operation and a second operation and nine computation results are output through the matching line according to conditions of the ternary weight and ternary input.Type: ApplicationFiled: April 20, 2023Publication date: February 1, 2024Applicant: Inha University Research and Business FoundationInventors: Yeongkyo SEO, Dae Woong KWON
-
Publication number: 20240037382Abstract: The method includes: acquiring a plurality of to-be-treated optical signals with unequal wavelengths; inputting the to-be-treated optical signals into a micro-ring-resonator array, wherein the micro-ring-resonator array includes a plurality of micro-ring resonators that are connected in series; applying a corresponding electric current to the micro-ring-resonator array, to adjust a transfer function of each of the micro-ring resonators to reach a target value; and feeding an optical signal outputted by the micro-ring-resonator array into a photodiode, to obtain an operation result of the average pooling of the neural network.Type: ApplicationFiled: December 30, 2021Publication date: February 1, 2024Inventors: Jingjing CHEN, Ping HUANG, Ruizhen WU, Lin WANG
-
Publication number: 20240037383Abstract: Herein are machine learning (ML) explainability (MLX) techniques for calculating and using a novel fidelity metric for assessing and comparing explainers that are based on feature attribution. In an embodiment, a computer generates many anomalous tuples from many non-anomalous tuples. Each anomalous tuple contains a perturbed value of a respective perturbed feature. For each anomalous tuple, a respective explanation is generated that identifies a respective identified feature as a cause of the anomalous tuple being anomalous. A fidelity metric is calculated by counting correct explanations for the anomalous tuples whose identified feature is the perturbed feature. Tuples may represent entries in an activity log such as structured query language (SQL) statements in a console output log of a database server. This approach herein may gauge the quality of a set of MLX explanations for why log entries or network packets are characterized as anomalous by an intrusion detector or other anomaly detector.Type: ApplicationFiled: July 26, 2022Publication date: February 1, 2024Inventors: Kenyu Kobayashi, Arno Schneuwly, Renata Khasanova, Matteo Casserini, Felix Schmidt
-
Publication number: 20240037384Abstract: Systems and methods are provided for updating data in a computer network. An exemplary method includes: receiving input data from at least one device; performing an extraction operation on the input data to extract at least one feature; producing at least one feature vector based on the at least one feature; performing a similarity analysis between the at least one feature vector and a plurality of other feature vectors from a plurality of autoencoders; selecting a first autoencoder from the plurality of autoencoders demonstrating significant similarity with at least one feature vector; determining whether the input data exhibits a recurring drift or a new drift; and training a new autoencoder using at least a portion of the input data.Type: ApplicationFiled: July 27, 2022Publication date: February 1, 2024Applicant: Raytheon CompanyInventors: Kin Gwn Lore, Philip A. Sallee, Franklin R. Tanner
-
Publication number: 20240037385Abstract: Embodiments described herein include a process for detecting energy tool activations. The process can begin by receiving a surgical video of a surgical procedure involving energy tool activations. The process then applies a sequence of sampling windows to the surgical video to generate a sequence of windowed samples of the surgical video. Next, for each windowed sample in the sequence of windowed samples, the process applies a deep-learning model to a sequence of video frames within the windowed sample to generate an activation/non-activation inference and a confidence level associated with the activation/non-activation inference for the windowed sample. As a result, a sequence of activation/non-activation inferences and a sequence of associated confidence levels are generated. The process subsequently identifies a sequence of activation events in the surgical video based on the sequence of activation/non-activation inferences and the sequence of associated confidence levels.Type: ApplicationFiled: July 27, 2022Publication date: February 1, 2024Inventors: Meysam TORABI, Varun GOEL, Jocelyn BARKER, Rami ABUKHALIL, Richard W. TIMM, Pablo E. GARCIA KILROY
-
Publication number: 20240037386Abstract: A method, system, apparatus, and non-transitory computer-readable medium for image processing using a multi-task neural network framework may be provided. The method be executed by one or more processors, and may include receiving an input image; and generating one or more image patches based on the input image. The method may include performing an image processing task based on the input image using the multi-task neural network framework, wherein the multi-task neural network framework is trained using an adaptive feature distillation function, and wherein the adaptive feature distillation function is based on a comparison of intermediate features of the multi-task neural network framework and intermediate features of a plurality of single-task neural network models. The method may include generating an output of the image processing task based on up sampling an output of the multi-task neural network framework.Type: ApplicationFiled: July 28, 2022Publication date: February 1, 2024Applicant: Rakuten Group, Inc.Inventors: Geethu JACOB, Vishal AGARWAL, Bjorn STENGER
-
Publication number: 20240037387Abstract: A power transformer fault diagnosis method based on a stacked time series network, includes: collecting gas-in-oil data of a transformer in each substation; performing z-score normalization on the collected data to obtain a normalized matrix; dividing the normalized matrix into a training set and a test set in proportion; constructing a stacked time series network based on Xgboost and a bidirectional gated neural network, and inputting the training set and the test set to perform network training; and normalizing real-time collected data to obtain trainable data to predict a fault and update network parameters. The gas-in-oil data is predicted by using Xgboost and a gated neural network, obtains prediction data of a power transformer from two time series networks by using a meta learner, and obtains a fault diagnosis result of the transformer by using a Softmax layer. The neural network has accurate fault diagnosis performance and stable robustness.Type: ApplicationFiled: December 1, 2022Publication date: February 1, 2024Applicants: WUHAN UNIVERSITY, State Grid Tianjin Electric Power CompanyInventors: Yigang HE, Zhikai XING, Xiao WANG, Xiaoyu LIU, Xue JIANG, Qingwu GONG, Jianfeng WANG, Shiqian MA
-
Publication number: 20240037388Abstract: A method of learning a neural network, wherein the neural network includes: a feature selection layer for selecting a part of input data; a feature extraction layer for extracting a feature quantity on the basis of the selected input data; a prediction layer for performing a prediction on the basis of the feature quantity; and a partial reconstruction layer for reconstructing the selected input data on the basis of the feature quantity, and the method includes adjusting a weight parameter of the neural network on the basis of a prediction accuracy by the prediction layer and a reconstruction error in the partial reconstruction layer.Type: ApplicationFiled: July 25, 2023Publication date: February 1, 2024Applicant: NEC CorporationInventor: Masanao Natsumeda
-
Publication number: 20240037389Abstract: A method of learning a neural network, wherein the neural network includes: a feature selection layer for selecting a part of input data including information about a domain of each sample; a feature extraction layer for extracting a feature quantity on the basis of the selected input data; and a prediction layer for performing a prediction on the basis of the feature quantity, and the method includes adjusting a weight parameter of the neural network to increase a prediction accuracy by the prediction layer and to reduce a contribution to a prediction result of the prediction layer by the domain of the input data.Type: ApplicationFiled: July 27, 2023Publication date: February 1, 2024Applicant: NEC CorporationInventor: Masanao NATSUMEDA
-
Publication number: 20240037390Abstract: A method for training a weight-sharing neural network with stochastic architectures is disclosed. The method includes (i) selecting a mini-batch from a plurality of mini-batches, a training data set for a task being grouped into the plurality of mini-batches and each of the plurality of mini-batches comprising a plurality of instances: (ii) stochastically selecting a plurality of network architectures of the neural network for the selected mini-batch; (iii) obtaining a loss for each instance of the selected mini-batch by applying the instance to one of the plurality of network architectures; and (iv) updating shared weights of the neural network based on the loss for each instance of the selected mini-batch.Type: ApplicationFiled: October 15, 2020Publication date: February 1, 2024Inventors: Jun Zhu, Zhijie Deng, Yinpeng Dong, Chao Zhang, Kevin Yang
-
Publication number: 20240037391Abstract: A method for distributed training of a graph-embedding neural network is disclosed. The method, performed at a first server, includes computing, based on a first input data sample, first model data and first embedding data of a first graph neural network, the first graph neural network corresponding to a first set of nodes of a graph that are visible to the first server, and includes sharing the first model data and the first embedding data with a second server. The method also includes receiving second embedding data from a third server, the second embedding data comprising embedding data of a second graph neural network corresponding to a second set of nodes of the graph that are invisible to the first server, and includes computing second model data of the first graph neural network based on a second input data sample and the embedding data of the second graph neural network.Type: ApplicationFiled: December 15, 2021Publication date: February 1, 2024Inventors: Lan Wang, Lei Yu, Li Jiang
-
Publication number: 20240037392Abstract: A method for further training of a neural network for processing measurement data, which neural network has been pre-trained with training examples from a set M. In the method: a batch B of new training examples is provided; a subset D?M of the previous training examples is provided; the new training examples from batch B and the previous training examples from subset D are processed by the neural network into outputs respectively; the deviations of the outputs from the respective target outputs are evaluated using a predefined cost function; parameters characterizing the behavior of the neural network are optimized with the aim that, during further processing of previous and new training examples, the evaluation with the cost function is improved in regard to new training examples from batch B and is not made worse in regard to previous training examples from subset D.Type: ApplicationFiled: July 24, 2023Publication date: February 1, 2024Inventor: Frank Schmidt
-
Publication number: 20240037393Abstract: A method for training a control policy for controlling a technical system. The method includes training a neural network to implement a value function by: adapting the neural network for reducing a loss which, for a plurality of states and, for each state, for at least one action that has been previously carried out in the state, involves a deviation between a prediction for a cumulative reward and an estimation of the cumulative reward that is ascertained from a subsequent state that has been achieved by the action, and a reward that is obtained by the action. In the loss, for each action, the deviation for the action is weighted more strongly the greater the likelihood is that the action is selected by the control policy, in relation to the likelihood that the action is selected by a behavior control policy. The method also includes training the control policy.Type: ApplicationFiled: July 27, 2023Publication date: February 1, 2024Inventors: Fabian Otto, Gerhard Neumann, Anh Vien Ngo, Hanna Ziesche
-
Publication number: 20240037394Abstract: A neural network accelerator architecture for multiple task adaptation comprises a volatile memory comprising a plurality of subarrays, each subarray comprising M rows and N columns of volatile memory cells; a source line driver connected to a plurality of N source lines, each source line corresponding to a column in the subarray; a binary mask buffer memory having size at least N bits, each bit corresponding to a column in the subarray, where a 0 corresponds to turning off the column for a convolution operation and a 1 corresponds to turning on the column for the convolution operation; and a controller configured to selectively drive each of the N source lines with a corresponding value from the mask buffer; wherein each column in the subarray is configured to store a convolution kernel.Type: ApplicationFiled: July 27, 2023Publication date: February 1, 2024Applicant: Arizona Board of Regents on behalf of Arizona State UniversityInventors: Deliang Fan, Fan Zhang, Li Yang
-
Publication number: 20240037395Abstract: Provided herein are exemplary systems and methods including the generation of a superior strategy for deployment to real time actual conditions with dynamic feedback to the secure intelligent networked architecture in order for adjustments to be made to the strategy being deployed to the real time actual conditions and the learned generation of subsequent strategies.Type: ApplicationFiled: October 2, 2023Publication date: February 1, 2024Inventors: Howard M. Getson, Sean Vallie, Robert Jump, Wincenty Borodziewicz
-
Publication number: 20240037396Abstract: Computer systems and computer-implemented methods modify a machine learning network, such as a deep neural network, to introduce judgment to the network. A “combining” node is added to the network, to thereby generate a modified network, where activation of the combining node is based, at least in part, on output from a subject node of the network. The computer system then trains the modified network by, for each training data item in a set of training data, performing forward and back propagation computations through the modified network, where the backward propagation computation through the modified network comprises computing estimated partial derivatives of an error function of an objective for the network, except that the combining node selectively blocks back-propagation of estimated partial derivatives to the subject node, even though activation of the combining node is based on the activation of the subject node.Type: ApplicationFiled: October 9, 2023Publication date: February 1, 2024Applicant: D5AI LLCInventor: James K. Baker
-
Publication number: 20240037397Abstract: A method interprets a convolutional sequence model. The method converts an input data sequence having input segments into output features. The method clusters the input segments into clusters using respective resolution-controllable class prototypes allocated to each of classes. Each respective class prototype includes a respective output feature subset characterizing a respective associated class. The method calculates, using the clusters, similarity scores that indicate a similarity of an output feature to a respective class prototypes responsive to distances between the output feature and the respective class prototypes. The method concatenates the similarity scores to obtain a similarity vector. The method performs a prediction and prediction support operation that provides a value of prediction and an interpretation for the value responsive to the input segments and similarity vector.Type: ApplicationFiled: October 2, 2023Publication date: February 1, 2024Inventors: Jingchao Ni, Zhengzhang Chen, Wei Cheng, Bo Zong, Haifeng Chen