公开(公告)号：CA2836104C

公开(公告)日：2017-04-11

申请号：CA2836104

申请日：2012-05-21

Applicant: QUALCOMM INC

Inventor： GROKOP LEONARD H ,  SARAH ANTHONY

IPC: G06F3/01 ,  G06F3/033

Abstract: Techniques are described herein for classifying multiple device states using separate Bayesian classifiers. An example of a method described herein includes accessing sensor information of a device, wherein at least some of the sensor information is used in a first feature set and at least some of the sensor information is used in a second feature set; processing the first feature set using a first classification algorithm configured to determine a first proposed state of a first state type and a first proposed state of a second state type; processing the second feature set using a second classification algorithm configured to determine a second proposed state of the first state type and a second proposed state of the second state type; and determining a proposed state of the device as the first proposed state of the first state type and the second proposed state of the second state type.

公开(公告)号：CA2924048A1

公开(公告)日：2015-04-09

申请号：CA2924048

申请日：2014-08-21

Applicant: QUALCOMM INC

Inventor： SARAH ANTHONY ,  KIMBALL ROBERT HOWARD

IPC: G06N3/02 ,  G06N3/08

Abstract: A method for improving neural dynamics includes obtaining prototypical neuron dynamics. The method also includes modifying parameters of a neuron model so that the neuron model matches the prototypical neuron dynamics. The neuron dynamics comprise membrane voltages and/or spike timing.

公开(公告)号：BR112013030271A2

公开(公告)日：2016-11-29

申请号：BR112013030271

申请日：2012-05-21

Applicant: QUALCOMM INC

Inventor： SARAH ANTHONY ,  GROKOP LEONARD

IPC: G06F3/033 ,  G06F3/01

Abstract: resumo “método e aparelho para classificar múltiplos estados de dispositivo” trata-se de técnicas descritas no presente documento para classificar múltiplos estados de dispositivo com o uso de classificadores bayesianos separados. um exemplo de um método descrito no presente documento inclui acessar informações de sensor de um dispositivo, em que pelo menos algumas das informações de sensor são usadas em um primeiro conjunto de recursos e pelo menos algumas das informações de sensor são usadas em um segundo conjunto de recursos; processar o primeiro conjunto de recursos com o uso de um primeiro algoritmo de classificação configurado para determinar um primeiro estado proposto de um primeiro tipo de estado e um primeiro estado proposto de um segundo tipo de estado; processar o segundo conjunto de recursos com o uso de um segundo algoritmo de classificação configurado para determinar um segundo estado proposto do primeiro tipo de estado e um segundo estado proposto do segundo tipo de estado; e determinar um estado proposto do dispositivo como o primeiro estado proposto do primeiro tipo de estado e o segundo estado proposto do segundo tipo de estado. 1/1

公开(公告)号：CA2937945A1

公开(公告)日：2015-09-03

申请号：CA2937945

申请日：2015-02-12

Applicant: QUALCOMM INC

Inventor： SARAH ANTHONY ,  LEVIN JEFFREY ALEXANDER ,  GEHLHAAR JEFFREY BAGINSKY

IPC: G06N3/04 ,  G06N3/10

Abstract: Certain aspects of the present disclosure support efficient implementation of common neuron models. In an aspect, a first memory layout can be allocated for parameters and state variables of instances of a first neuron model, and a second memory layout different from the first memory layout can be allocated for parameters and state variables of instances of a second neuron model having a different complexity than the first neuron model.

公开(公告)号：CA2926034A1

公开(公告)日：2015-04-23

申请号：CA2926034

申请日：2014-08-21

Applicant: QUALCOMM INC

Inventor： SARAH ANTHONY ,  KIMBALL ROBERT HOWARD ,  SPINAR BRIAN

IPC: G06N3/04

Abstract: A method for dynamically modifying synaptic delays in a neural network includes initializing a delay parameter and operating the neural network. The method further includes dynamically updating the delay parameter based on a program which is based on a statement including the delay parameter.

公开(公告)号：CA2836104A1

公开(公告)日：2012-12-06

申请号：CA2836104

申请日：2012-05-21

Applicant: QUALCOMM INC

Inventor： GROKOP LEONARD H ,  SARAH ANTHONY

IPC: G06F3/01 ,  G06F3/033

Abstract: Techniques are described herein for classifying multiple device states using separate Bayesian classifiers. An example of a method described herein includes accessing sensor information of a device, wherein at least some of the sensor information is used in a first feature set and at least some of the sensor information is used in a second feature set; processing the first feature set using a first classification algorithm configured to determine a first proposed state of a first state type and a first proposed state of a second state type; processing the second feature set using a second classification algorithm configured to determine a second proposed state of the first state type and a second proposed state of the second state type; and determining a proposed state of the device as the first proposed state of the first state type and the second proposed state of the second state type.

公开(公告)号：WO2015148190A3

公开(公告)日：2015-12-10

申请号：PCT/US2015021092

申请日：2015-03-17

Applicant: QUALCOMM INC

Inventor： ANNAPUREDDY VENKATA SREEKANTA REDDY ,  JULIAN DAVID JONATHAN ,  SARAH ANTHONY

IPC: G06N3/04 ,  G06N3/08

CPC classification number: G06N3/0454 ,  G06N3/049 ,  G06N3/088

Abstract: A method of distributed computation includes computing a first set of results in a first computational chain with a first population of processing nodes and passing the first set of results to a second population of processing nodes. The method also includes entering a first rest state with the first population of processing nodes after passing the first set of results and computing a second set of results in the first computational chain with the second population of processing nodes based on the first set of results. The method further includes passing the second set of results to the first population of processing nodes, entering a second rest state with the second population of processing nodes after passing the second set of results and orchestrating the first computational chain.

Abstract translation: 一种分布式计算的方法包括：在具有第一种处理节点的第一计算链中计算第一组结果，并将第一组结果传递给第二种处理节点。 该方法还包括在传递第一组结果之后，与处理节点的第一群体一起输入第一休息状态，并且基于第一组结果与第二种处理节点计算第一计算链中的第二组结果。 该方法还包括将第二组结果传递给第一处理节点群，在通过第二组结果并编排第一计算链之后，与第二处理节点群进入第二休息状态。

公开(公告)号：EP3295383A1

公开(公告)日：2018-03-21

申请号：EP16719637

申请日：2016-04-14

Applicant: QUALCOMM INC

Inventor： LIN DEXU ,  BADIN MATTHEW ,  HOWARD DAVID EDWARD ,  DIJKMAN DANIEL HENDRICUS FRANCISCUS ,  TREMAINE MICHAEL COLIN ,  SARAH ANTHONY

IPC: G06N3/063

CPC classification number: G06N3/08 ,  G06N3/063 ,  G06N99/005

Abstract: A method of reducing computational complexity for a fixed point neural network operating in a system having a limited bit width in a multiplier-accumulator (MAC) includes reducing a number of bit shift operations when computing activations in the fixed point neural network. The method also includes balancing an amount of quantization error and an overflow error when computing activations in the fixed point neural network.

公开(公告)号：WO2017023551A3

公开(公告)日：2017-04-13

申请号：PCT/US2016043146

申请日：2016-07-20

Applicant: QUALCOMM INC

Inventor： MAJUMDAR SOMDEB ,  LIN DEXU ,  TOWAL REGAN BLYTHE ,  SARAH ANTHONY

IPC: G06N99/00 ,  G06K9/62 ,  G06N3/04

CPC classification number: G06N3/08 ,  G06K9/4628 ,  G06K9/6256 ,  G06K9/6269 ,  G06N3/0445 ,  G06N3/0454 ,  G06N99/005

Abstract: A method of detecting unknown classes is presented and includes generating a first classifier for multiple first classes. In one configuration, an output of the first classifier has a dimension of at least two. The method also includes designing a second classifier to receive the output of the first classifier to decide whether input data belongs to the multiple first classes or at least one second class.

Abstract translation: 提出了一种检测未知类的方法，包括生成多个第一类的第一分类器。 在一个配置中，第一分类器的输出具有至少两个维度。 该方法还包括设计第二分类器以接收第一分类器的输出以确定输入数据是属于多个第一类还是至少一个第二类。

公开(公告)号：WO2015050641A3

公开(公告)日：2015-06-04

申请号：PCT/US2014052131

申请日：2014-08-21

Applicant: QUALCOMM INC

Inventor： SARAH ANTHONY ,  KIMBALL ROBERT HOWARD

IPC: G06N3/04

CPC classification number: G06N3/049 ,  G06N3/04 ,  G06N3/063 ,  G06N3/08 ,  G06N3/082 ,  G06N99/005

Abstract: A method for improving neural dynamics includes obtaining prototypical neuron dynamics. The method also includes modifying parameters of a neuron model so that the neuron model matches the prototypical neuron dynamics. The neuron dynamics comprise membrane voltages and/or spike timing.

Abstract translation: 一种改善神经动力学的方法包括获得原型神经元动力学。 该方法还包括修改神经元模型的参数，使得神经元模型与原型神经元动力学匹配。 神经元动态包括膜电压和/或尖峰定时。