公开(公告)号：US20190251461A1

公开(公告)日：2019-08-15

申请号：US16277956

申请日：2019-02-15

申请人： BABYLON PARTNERS LIMITED

发明人： Laura Helen DOUGLAS ,  Pavel MYSHKOV ,  Robert WALECKI ,  Iliyan Radev ZAROV ,  Konstantinos GOURGOULIAS ,  Christopher LUCAS ,  Christopher Robert HART ,  Adam Philip BAKER ,  Maneesh SAHANI ,  Iurii PEROV ,  Saurabh JOHRI

IPC分类号： G06N7/00 ,  G06N3/08 ,  G06N20/20 ,  G06F17/16 ,  G16H50/30 ,  G16H50/20

CPC分类号： G16H50/30 ,  G06F17/16 ,  G06N3/0454 ,  G06N3/08 ,  G06N3/082 ,  G06N5/04 ,  G06N7/005 ,  G06N20/20 ,  G16H50/20 ,  G16H50/70

摘要： Methods for providing a computer implemented medical diagnosis are provided. In one aspect, a method includes receiving an input from a user comprising at least one symptom of the user, and providing the at least one symptom as an input to a medical model. The method also includes deriving estimates of the probability of the user having a disease from the discriminative model, inputting the estimates to the inference engine, performing approximate inference on the probabilistic graphical model to obtain a prediction of the probability that the user has that disease, and outputting the probability of the user having the disease for display by a display device.

公开(公告)号：US20190236354A1

公开(公告)日：2019-08-01

申请号：US16257160

申请日：2019-01-25

申请人： Panasonic Intellectual Property Corporation of America

发明人： Yohei NAKATA ,  Yasunori ISHII

IPC分类号： G06K9/00 ,  G06N5/04 ,  G06N3/08 ,  G06N20/00

CPC分类号： G06K9/00536 ,  G06N3/08 ,  G06N3/082 ,  G06N5/04 ,  G06N20/00

摘要： An information processing method including the following executed using a computer: obtaining a neural network model that solves a regression problem; obtaining input data and label data corresponding to the input data; compressing a network of the neural network model to obtain a compressed model; and changing the label data and the number of nodes in the neural network model, based on information indicating performance of the compressed model, the number of nodes being assigned to the regression problem, the information being calculated using the label data and output data which is obtained by inputting the input data to the compressed model.

公开(公告)号：US20190197406A1

公开(公告)日：2019-06-27

申请号：US15853458

申请日：2017-12-22

申请人： Microsoft Technology Licensing, LLC

发明人： Bita Darvish Rouhani ,  Douglas C. Burger ,  Eric S. Chung

IPC分类号： G06N3/08 ,  G06F15/18

CPC分类号： G06N3/082 ,  G06N3/063 ,  G06N20/00

摘要： A computer implemented method of optimizing a neural network includes obtaining a deep neural network (DNN) trained with a training dataset, determining a spreading signal between neurons in multiple adjacent layers of the DNN wherein the spreading signal is an element-wise multiplication of input activations between the neurons in a first layer to neurons in a second next layer with a corresponding weight matrix of connections between such neurons, and determining neural entropies of respective connections between neurons by calculating an exponent of a volume of an area covered by the spreading signal. The DNN may be optimized based on the determined neural entropies between the neurons in the multiple adjacent layers.

公开(公告)号：US20190188554A1

公开(公告)日：2019-06-20

申请号：US16283021

申请日：2019-02-22

申请人： Intel Corporation

发明人： Liwei Ma ,  Elmoustapha Ould-Ahmed-Vall ,  Barath Lakshmanan ,  Ben J. Ashbaugh ,  Jingyi Jin ,  Jeremy Bottleson ,  Mike B. Macpherson ,  Kevin Nealis ,  Dhawal Srivastava ,  Joydeep Ray ,  Ping T. Tang ,  Michael S. Strickland ,  Xiaoming Chen ,  Anbang Yao ,  Tatiana Shpeisman ,  Altug Koker ,  Abhishek R. Appu

IPC分类号： G06N3/04 ,  G06N3/08 ,  G06T1/20 ,  G06N3/063

CPC分类号： G06N3/04 ,  G06N3/0445 ,  G06N3/0454 ,  G06N3/063 ,  G06N3/082 ,  G06T1/20

摘要： Embodiments provide systems and methods which facilitate optimization of a convolutional neural network (CNN). One embodiment provides for a non-transitory machine-readable medium storing instructions that cause one or more processors to perform operations comprising processing a trained convolutional neural network (CNN) to generate a processed CNN, the trained CNN having weights in a floating-point format. Processing the trained CNN includes quantizing the weights in the floating-point format to generate weights in an integer format. Quantizing the weights includes generating a quantization table to enable non-uniform quantization of the weights and quantizing the weights from the floating-point format to the integer format using the quantization table. The operations additionally comprise performing an inference operation utilizing the processed CNN with the integer format weights.

公开(公告)号：US20190187718A1

公开(公告)日：2019-06-20

申请号：US15844732

申请日：2017-12-18

申请人： GM GLOBAL TECHNOLOGY OPERATIONS LLC

发明人： Guangyu J. Zou

IPC分类号： G05D1/02 ,  G06K9/00 ,  G06K9/66 ,  G06T5/20 ,  G06T1/20

CPC分类号： G05D1/0231 ,  G06K9/00798 ,  G06K9/66 ,  G06N3/02 ,  G06N3/0454 ,  G06N3/082 ,  G06T1/20 ,  G06T5/20

摘要： Systems and methods are provided for detecting features from multi-modal image-like data representations. The system includes a wavelet transformer configured to, via at least one processor, receive image data and to wavelet transform the image data, thereby providing decomposed image data divided into frequency sub-bands. The system further includes an artificial neural network configured to receive and process at least one sub-band of the decomposed image data to detect image features based thereon, the artificial neural network configured to output the detected image features.

公开(公告)号：US20190005357A1

公开(公告)日：2019-01-03

申请号：US15635367

申请日：2017-06-28

申请人： Applied Materials, Inc.

发明人： Sreekar BHAVIRIPUDI ,  Shreekant GAYAKA

IPC分类号： G06K9/62 ,  G06N3/08 ,  G06N3/04

CPC分类号： G06K9/6267 ,  G06K9/4628 ,  G06K9/6256 ,  G06K9/6271 ,  G06N3/04 ,  G06N3/0454 ,  G06N3/08 ,  G06N3/082 ,  G06N20/00

摘要： A method of classifying substrates with a metrology tool is herein disclosed. The method begins by training a deep learning framework using convolutional neural networks with a training dataset for classifying image dataset. Obtaining a new image from the meteorology tool. Running the new image through the deep learning framework to classify the new image.

公开(公告)号：US20180261214A1

公开(公告)日：2018-09-13

申请号：US15848199

申请日：2017-12-20

申请人： Facebook, Inc.

发明人： Jonas Gehring ,  Michael Auli ,  Yann Nicolas Dauphin ,  David G. Grangier ,  Dzianis Yarats

IPC分类号： G10L15/16 ,  G06N3/04 ,  G06N3/08 ,  G10L15/22 ,  G06F17/28

CPC分类号： G10L15/16 ,  G06F17/2836 ,  G06N3/04 ,  G06N3/0445 ,  G06N3/0454 ,  G06N3/08 ,  G06N3/082 ,  G06N3/084 ,  G10L15/22

摘要： Exemplary embodiments relate to improvements to neural networks for translation and other sequence-to-sequence tasks. A convolutional neural network may include multiple blocks, each having a convolution layer and gated linear units; gating may determine what information passes through to the next block level. Residual connections, which add the input of a block back to its output, may be applied around each block. Further, an attention may be applied to determine which word is most relevant to translate next. By applying repeated passes of the attention to multiple layers of the decoder, the decoder is able to work on the entire structure of a sentence at once (with no temporal dependency). In addition to better accuracy, this configuration is better at capturing long-range dependencies, better models the hierarchical syntax structure of a sentence, and is highly parallelizable and thus faster to run on hardware.

公开(公告)号：US20180259608A1

公开(公告)日：2018-09-13

申请号：US15779448

申请日：2016-11-29

申请人： ARTERYS INC.

发明人： Daniel Irving Golden ,  John Axerio-Cilies ,  Matthieu Le ,  Torin Arni Taerum ,  Jesse Lieman-Sifry

IPC分类号： G01R33/56 ,  G06T7/11 ,  G06N3/02 ,  G01R33/563

CPC分类号： G01R33/5608 ,  G01R33/561 ,  G01R33/5613 ,  G01R33/56308 ,  G01R33/56316 ,  G06N3/006 ,  G06N3/02 ,  G06N3/0454 ,  G06N3/082 ,  G06N3/084 ,  G06T7/0012 ,  G06T7/11 ,  G06T2207/10088 ,  G06T2207/10132 ,  G06T2207/20081 ,  G06T2207/20084 ,  G06T2207/30048

摘要： Systems and methods for automated segmentation of anatomical structures, such as the human heart. The systems and methods employ convolutional neural networks (CNNs) to autonomously segment various parts of an anatomical structure represented by image data, such as 3D MRI data. The convolutional neural network utilizes two paths, a contracting path which includes convolution/pooling layers, and an expanding path which includes upsampling/convolution layers. The loss function used to validate the CNN model may specifically account for missing data, which allows for use of a larger training set. The CNN model may utilize multi-dimensional kernels (e.g., 2D, 3D, 4D, 6D), and may include various channels which encode spatial data, time data, flow data, etc. The systems and methods of the present disclosure also utilize CNNs to provide automated detection and display of landmarks in images of anatomical structures.

公开(公告)号：US20180232640A1

公开(公告)日：2018-08-16

申请号：US15488430

申请日：2017-04-14

申请人： Samsung Electronics Co., Ltd.

发明人： Zhengping JI ,  John Wakefield BROTHERS ,  Ilia OVSIANNIKOV ,  Eunsoo SHIM

IPC分类号： G06N3/08

CPC分类号： G06N3/082

摘要： An embodiment includes a method, comprising: pruning a layer of a neural network having multiple layers using a threshold; and repeating the pruning of the layer of the neural network using a different threshold until a pruning error of the pruned layer reaches a pruning error allowance.

公开(公告)号：US20180210939A1

公开(公告)日：2018-07-26

申请号：US15880225

申请日：2018-01-25

申请人： HRL Laboratories, LLC

发明人： Youngkwan Cho ,  Hyukseong Kwon ,  Rajan Bhattacharyya

IPC分类号： G06F17/30 ,  G06N3/08

CPC分类号： G06F16/28 ,  G06F16/137 ,  G06F16/2477 ,  G06N3/082 ,  G06N5/022 ,  G06N20/00

摘要： Described is a system for an episodic memory used by an automated platform. The system acquires data from an episodic memory that comprises an event database, an event-sequence graph, and an episode list. Using the event-sequence graph, the system identifies a closest node to a current environment for the automated platform. Based on the closest node and using a hash function or key based on the hash function, the system retrieves from the event database an episode that corresponds to the closest node, the episode including a sequence of events. Behavior of the automated platform in the current environment is guided based on the data from the episodic memory.