Patents Examined by Alan Chen
  • Patent number: 11455514
    Abstract: A method for determining a placement for machine learning model operations across multiple hardware devices includes receiving data specifying machine learning operations, and determining a placement that assigns each of the operations specified by the data to a respective device from the multiple hardware devices.
    Type: Grant
    Filed: August 28, 2019
    Date of Patent: September 27, 2022
    Assignee: Google LLC
    Inventors: Benoit Steiner, Anna Darling Goldie, Jeffrey Adgate Dean, Hieu Hy Pham, Azalia Mirhoseini, Quoc V. Le
  • Patent number: 11451565
    Abstract: Techniques are provided herein for contextual embedding of features of operational logs or network traffic for anomaly detection based on sequence prediction. In an embodiment, a computer has a predictive recurrent neural network (RNN) that detects an anomalous network flow. In an embodiment, an RNN contextually transcodes sparse feature vectors that represent log messages into dense feature vectors that may be predictive or used to generate predictive vectors. In an embodiment, graph embedding improves feature embedding of log traces. In an embodiment, a computer detects and feature-encodes independent traces from related log messages. These techniques may detect malicious activity by anomaly analysis of context-aware feature embeddings of network packet flows, log messages, and/or log traces.
    Type: Grant
    Filed: September 5, 2018
    Date of Patent: September 20, 2022
    Assignee: Oracle International Corporation
    Inventors: Guang-Tong Zhou, Hossein Hajimirsadeghi, Andrew Brownsword, Stuart Wray, Craig Schelp, Rod Reddekopp, Felix Schmidt
  • Patent number: 11437143
    Abstract: Various systems and methods for predicting metabolic and bariatric surgery outcomes are provided. The systems and methods can also provide predictions for non-surgical metabolic and bariatric treatments. In general, a user can receive predictive outcomes of multiple bariatric procedures that could be performed on a patient. In one embodiment, a user can electronically access a metabolic and bariatric surgery outcome prediction system, e.g., using one or more web pages. The system can provide predictive outcomes of one or more different bariatric surgeries for the patient based on data gathered from the user and on historical data regarding outcomes of the different bariatric surgeries. The system can additionally provide predictive outcomes for not having any treatment and/or a comparison of the predictive outcomes of the one or more different bariatric surgeries to the predictive outcomes for not having any treatment.
    Type: Grant
    Filed: January 28, 2019
    Date of Patent: September 6, 2022
    Assignee: Ethicon Endo-Surgery, Inc.
    Inventors: Jason L. Harris, Christopher J. Hess, Nitin Kumar Jain, Diane M. Francis, Thomas E. Albrecht, Tina Denise Hunter
  • Patent number: 11436537
    Abstract: A method can include determining a cell of a grid to which a first feature and a second feature of each of a plurality of input/output examples maps, determining an average of respective features of the cell to generate respective level 2 synthetic feature vectors, for each cell with an input/output example of the input/output examples mapped thereto, generating a sub-grid of cells and map the input/output examples mapped to a cell of the sub-grid, determining an average of respective features to generate respective level 1 synthetic feature vectors comprising the average of the respective features, training the ML technique using the level 2 synthetic feature vector, testing the trained ML technique using the level 1 synthetic feature vector of each sub-cell, and further testing the trained ML technique using the input/output examples to generate a class and confidence for each of the input/output examples.
    Type: Grant
    Filed: March 8, 2019
    Date of Patent: September 6, 2022
    Assignee: Raytheon Company
    Inventors: Holger M. Jaenisch, James W. Handley, Guy G. Swope
  • Patent number: 11423301
    Abstract: Methods, systems, and apparatus, including computer programs encoded on computer storage media, for training and deploying machine-learned communication over radio frequency (RF) channels. One method includes: determining an encoder and a decoder, at least one of which is configured to implement an encoding or decoding that is based on at least one of an encoder machine-learning network or a decoder machine-learning network that has been trained to encode or decode information over a communication channel; determining first information; using the encoder to process the first information and generate a first RF signal; transmitting, by at least one transmitter, the first RF signal through the communication channel; receiving, by at least one receiver, a second RF signal that represents the first RF signal altered by transmission through the communication channel; and using the decoder to process the second RF signal and generate second information as a reconstruction of the first information.
    Type: Grant
    Filed: February 21, 2019
    Date of Patent: August 23, 2022
    Assignee: Virginia Tech Intellectual Properties, Inc.
    Inventor: Timothy James O'Shea
  • Patent number: 11417179
    Abstract: A method of tracking a user in a store and tracking takes and puts in regard to items in said store. One method includes tracking a shopper in a physical store using two or more cameras that are overlapping to infer and account for shopping activity performed by the shopper. The method includes providing output of at least one of the two or more cameras to a processing entity to extract skeletal limb features of the shopper. Then, processing the skeletal limb features of the shopper to detect a take of an item from the store into possession of the shopper. The method further includes detecting, based on the tracking, that the shopper has exited the store and processing a charge to an account of shopper for the item based on the detected take.
    Type: Grant
    Filed: December 29, 2018
    Date of Patent: August 16, 2022
    Inventors: Gary M. Zalewski, Albert S. Penilla
  • Patent number: 11416737
    Abstract: A neural processing unit (NPU), a method for driving an artificial neural network (ANN) model, and an ANN driving apparatus are provided. The NPU includes a semiconductor circuit that includes at least one processing element (PE) configured to process an operation of an artificial neural network (ANN) model; and at least one memory configurable to store a first kernel and a first kernel filter. The NPU is configured to generate a first modulation kernel based on the first kernel and the first kernel filter and to generate second modulation kernel based on the first kernel and a second kernel filter generated by applying a mathematical function to the first kernel filter. Power consumption and memory read time are both reduced by decreasing the data size of a kernel read from a separate memory to an artificial neural network processor and/or by decreasing the number of memory read requests.
    Type: Grant
    Filed: October 12, 2021
    Date of Patent: August 16, 2022
    Assignee: DEEPX CO., LTD.
    Inventor: Lok Won Kim
  • Patent number: 11410047
    Abstract: Systems and methods for anomaly detection includes accessing first data comprising a plurality of historical reversion transactions. A plurality of legitimate transactions are determined from the plurality of historical reversion transactions. An autoencoder is trained using the plurality of legitimate transactions to generate a trained autoencoder capable of measuring a given transaction for similarity to the plurality of legitimate transactions. A first reconstructed transaction is generated by the trained autoencoder using a first transaction. The first transaction is determined to be anomalous based on a reconstruction difference between the first transaction and the first reconstructed transaction.
    Type: Grant
    Filed: December 31, 2018
    Date of Patent: August 9, 2022
    Assignee: PAYPAL, INC.
    Inventors: Liron Florens Ben Kimon, Michael Dymshits, Albert Zelmanovitch, Dan Ayash
  • Patent number: 11403528
    Abstract: A method of compressing a pre-trained deep neural network model includes inputting the pre-trained deep neural network model as a candidate model. The candidate model is compressed by increasing sparsity of the candidate, removing at least one batch normalization layer present in the candidate model, and quantizing all remaining weights into fixed-point representation to form a compressed model. Accuracy of the compressed model is then determined utilizing an end-user training and validation data set. Compression of the candidate model is repeated when the accuracy improves. Hyper parameters for compressing the candidate model are adjusted, then compression of the candidate model is repeated when the accuracy declines. The compressed model is output for inference utilization when the accuracy meets or exceeds the end-user performance metric and target.
    Type: Grant
    Filed: April 18, 2019
    Date of Patent: August 2, 2022
    Assignee: Kneron (Taiwan) Co., Ltd.
    Inventors: Bike Xie, Junjie Su, Jie Wu, Bodong Zhang, Chun-Chen Liu
  • Patent number: 11403527
    Abstract: A computing device trains a neural network machine learning model. A forward propagation of a first neural network is executed. A backward propagation of the first neural network is executed from a last layer to a last convolution layer to compute a gradient vector. A discriminative localization map is computed for each observation vector with the computed gradient vector using a discriminative localization map function. An activation threshold value is selected for each observation vector from at least two different values based on a prediction error of the first neural network. A biased feature map is computed for each observation vector based on the activation threshold value selected for each observation vector. A masked observation vector is computed for each observation vector using the biased feature map. A forward and a backward propagation of a second neural network is executed a predefined number of iterations using the masked observation vector.
    Type: Grant
    Filed: October 13, 2021
    Date of Patent: August 2, 2022
    Assignee: SAS Institute Inc.
    Inventors: Xinmin Wu, Yingjian Wang, Xiangqian Hu
  • Patent number: 11399060
    Abstract: A method, computer program product, and computer system for applying deductive artificial intelligence (AI) attribution and auditability to data inputs, wherein the deductive AI may account for ontologies and competing system information, and wherein the deductive AI attribution and auditability may be applied to the data inputs by vendor workflow. The data inputs applied with the deductive AI attribution and auditability may be processed via a feedback loop to align a sense-understand-decide-act (SUDA) understanding with an inductive AI understanding. The inductive AI may be automated via the feedback loop based upon, at least in part, an AI expert system processing of the data inputs. One or more policy based rules may be developed for user automation authorization based upon, at least in part, the feedback loop.
    Type: Grant
    Filed: January 18, 2019
    Date of Patent: July 26, 2022
    Assignee: Phacil, LLC
    Inventor: Roger Joseph Morin
  • Patent number: 11386337
    Abstract: A method, system, and non-transitory compute readable medium for hidden evidence correlation and causation linking including a forecasting device configured to forecast hidden evidence found in relation to a user input in hidden cycle measurements into future forecasted cycle measurements, where the forecasted hidden cycles are transformed into an amplitude versus frequency histogram with each histogram being compared to each other histogram and determined if causation is a candidate, and if causation is a candidate, a probability density function is applied to produce a degree of causation of a causation link for the hidden evidence in relation to the user input.
    Type: Grant
    Filed: May 21, 2019
    Date of Patent: July 12, 2022
    Assignee: INTERNATIONAL BUSINESS MACHINES CORPORATION
    Inventors: Aaron Keith Baughman, Bridget Briana Beamon, Dong Chen, Peter Kenneth Malkin
  • Patent number: 11386543
    Abstract: The present invention relates to verification of damage to vehicles. More particularly, the present invention relates to a universal approach to automated generation of a damage estimate to a vehicle using images of the vehicle and verification of a manually-generated damage repair proposals using the automatically generated damage estimate. Aspects and/or embodiments seek to provide a computer-implemented method of generating one or more repair estimates from one or more photos of a damaged vehicle and comparing the generated estimate(s) to one or more input repair estimates to verify the one or more input repair estimates.
    Type: Grant
    Filed: May 19, 2021
    Date of Patent: July 12, 2022
    Assignee: Tractable Ltd
    Inventors: Razvan Ranca, Marcel Horstmann, Bjorn Mattsson, Janto Oellrich, Yih Kai Teh, Ken Chatfield, Franziska Kirschner, Rusen Aktas, Laurent Decamp, Mathieu Ayel, Julia Peyre, Shaun Trill, Crystal Van Oosterom
  • Patent number: 11379740
    Abstract: A cognitive learning method comprising: receiving data from a plurality of data sources; processing the data from the plurality of data sources to perform a cognitive learning operation, the processing being performed via a cognitive inference and learning system, the cognitive learning operation implementing a cognitive learning technique according to a cognitive learning framework, the cognitive learning operation applying the cognitive learning technique to generate a cognitive learning result; and, updating a destination based upon the learning result.
    Type: Grant
    Filed: October 7, 2019
    Date of Patent: July 5, 2022
    Assignee: Cognitive Scale, Inc.
    Inventors: Matthew Sanchez, Manoj Saxena
  • Patent number: 11379739
    Abstract: A cognitive learning method comprising receiving data from a plurality of data sources; processing the data from the plurality of data sources to perform a cognitive machine learning operation, the processing being performed via a cognitive inference and learning system, the cognitive learning operation implementing a cognitive learning technique according to a cognitive learning framework, the cognitive machine learning operation applying the cognitive learning technique via a machine learning algorithm to generate a cognitive learning result; and, updating a destination based upon the learning result.
    Type: Grant
    Filed: October 1, 2019
    Date of Patent: July 5, 2022
    Assignee: Cognitive Scale, Inc.
    Inventors: Matthew Sanchez, Manoj Saxena
  • Patent number: 11379724
    Abstract: Methods, systems, and apparatus, including computer programs encoded on computer storage media, for domain-specific pruning of neural networks are described. An exemplary method includes obtaining a first neural network trained based on a first training dataset; obtaining one or more second training datasets respectively from one or more domains; and training, based on the first neural network and the one or more second training datasets, a second neural network comprising the first neural network and one or more branches extended from the first neural network, wherein the second neural network is applicable for inferencing in the one or more domains, and the training comprises: training the one or more branches based respectively on the one or more second training datasets and an output of the first neural network.
    Type: Grant
    Filed: July 12, 2021
    Date of Patent: July 5, 2022
    Assignee: MOFFETT TECHNOLOGIES CO., LIMITED
    Inventors: Jiachao Liu, Enxu Yan
  • Patent number: 11379745
    Abstract: In an embodiment, a method includes: using at least two items of sequentially collected, mutually associated data to create at least two detection data sets, each including a first number of items of sequentially collected data in the at least two items of data; using an autoencoder to process the at least two detection data sets, to output result data sets respectively corresponding to the at least two detection data sets, the first number being equal to the number of neurons in an input layer of the autoencoder, and the autoencoder being trained using data having a regular pattern of variation identical to the at least two items of data; and determining, as abnormal data, data which does not have the regular pattern of variation in the at least two items of data, based upon the at least two detection data sets and the result data sets corresponding thereto.
    Type: Grant
    Filed: September 19, 2018
    Date of Patent: July 5, 2022
    Assignee: Siemens Aktiengesellschaft
    Inventors: Yi Yang, Jing Ma
  • Patent number: 11375030
    Abstract: One or more methods and/or techniques for providing a personalized future event notification to a user are provided herein. A content item (e.g., a news article, a social network post, etc.) may be evaluated utilizing a future event detection model to identify a future event (e.g., a festival). The future event detection model may have been trained to identify future events based upon part of speech analysis and entity recognition analysis of text within content items. In an example, the future event detection model may be used to identify locational features, temporal features, and/or entities from the content item. A user having a user interest in the future event above an interest threshold may be identified based upon user identifying information (e.g., a social network profile) being indicative of user interest in the future event. A personalized future event notification of the future event may be provided to the user.
    Type: Grant
    Filed: February 18, 2019
    Date of Patent: June 28, 2022
    Assignee: Yahoo Assets LLC
    Inventors: Junwei Jiang, Wenliang Cui, Qian Wan
  • Patent number: 11361225
    Abstract: A neural network architecture for attention-based efficient model adaptation is disclosed. A method includes accessing an input vector, the input vector comprising a numeric representation of an input to a neural network. The method includes providing the input vector to the neural network comprising a plurality of ordered layers, wherein each layer in at least a subset of the plurality of ordered layers is coupled with an adaptation module, wherein the adaptation module receives a same input value as a coupled layer for the adaptation module, and wherein an output value of the adaptation module is pointwise multiplied with an output value of the coupled layer to generate a next layer input value. The method includes generating an output of the neural network based on an output of a last one of the plurality of ordered layers in the neural network.
    Type: Grant
    Filed: December 18, 2018
    Date of Patent: June 14, 2022
    Assignee: Microsoft Technology Licensing, LLC
    Inventors: Mandar Dilip Dixit, Gang Hua
  • Patent number: 11361251
    Abstract: A computer system receives and stores data sets, a target metric, and a parameter that indicates a desired number of synthesized data sets. A hardware processor performs operations where each processing node of a neural network weights input data set values, determines gating operations to select processing operations, and generates a node output by applying the gating operations to weighted input data set values. The neural network is trained by modifying the gating operations, the input weight values, and the node output weight value until convergence. One or more nodes is selected having a larger magnitude node output weight value. Selected input data set values are processed with selected processing nodes using a selected subset of gating operations to produce the desired number of synthesized data sets. Names are generated for each of the synthesized data sets.
    Type: Grant
    Filed: January 16, 2020
    Date of Patent: June 14, 2022
    Assignee: Nasdaq, Inc.
    Inventor: Douglas Hamilton