公开(公告)号：WO2015156989A3

公开(公告)日：2015-12-03

申请号：PCT/US2015022024

申请日：2015-03-23

Applicant: QUALCOMM INC

Inventor： LEVIN JEFFREY ALEXANDER ,  LIU YINYIN ,  PENDYAM SANDEEP ,  CAMPOS MICHAEL

IPC: G06N3/04 ,  G06N99/00

CPC classification number: G06N99/005 ,  G06N3/049

Abstract: A method for maintaining a state variable in a synapse of a neural network includes maintaining a state variable in an axon. The state variable in the axon may be updated based on an occurrence of a first predetermined event. The method also includes updating the state variable in the synapse based on the state variable in the axon and an occurrence of a second predetermined event.

Abstract translation: 维持神经网络突触状态变量的方法包括维持轴突中的状态变量。 可以基于第一预定事件的发生来更新轴突中的状态变量。 该方法还包括基于轴突中的状态变量和第二预定事件的发生来更新突触中的状态变量。

公开(公告)号：WO2015127130A3

公开(公告)日：2015-12-10

申请号：PCT/US2015016693

申请日：2015-02-19

Applicant: QUALCOMM INC

Inventor： RAO NAVEEN GANDHAM ,  CAMPOS MICHAEL ,  LIU YINYIN

IPC: G06N3/04 ,  G06N3/02 ,  G06N3/08

CPC classification number: G06N3/08 ,  G06N3/049 ,  G06N3/088

Abstract: A method of dynamically modifying target selection with a neural network includes dynamically modifying a selection function by controlling an amount of imbalance of connections in the neural network. A selected neuron represents one of multiple candidate targets.

Abstract translation: 用神经网络动态修改目标选择的方法包括通过控制神经网络中的连接的不平衡量来动态修改选择功能。 选择的神经元代表多个候选目标之一。

公开(公告)号：WO2015023425A3

公开(公告)日：2015-05-14

申请号：PCT/US2014048605

申请日：2014-07-29

Applicant: QUALCOMM INC

Inventor： CANOY MICHAEL-DAVID NAKAYOSHI ,  LIU YINYIN ,  SARAH ANTHONY ,  MILNER ADRIENNE

IPC: G06N3/08 ,  G06N3/04

CPC classification number: G06N3/08 ,  G06N3/049

Abstract: Methods and apparatus are provided for training a neural device having an artificial nervous system by modulating at least one training parameter during the training. One example method for training a neural device having an artificial nervous system generally includes observing the neural device in a training environment and modulating at least one training parameter based at least in part on the observing. For example, the training apparatus described herein may modify the neural device's internal learning mechanisms (e.g., spike rate, learning rate, neuromodulators, sensor sensitivity, etc.) and/or the training environment's stimuli (e.g., move a flame closer to the device, make the scene darker, etc.). In this manner, the speed with which the neural device is trained (i.e., the training rate) may be significantly increased compared to conventional neural device training systems.

Abstract translation: 提供了用于通过在训练期间调制至少一个训练参数来训练具有人造神经系统的神经装置的方法和装置。 用于训练具有人造神经系统的神经装置的一个示例方法通常包括在训练环境中观察神经装置并且至少部分地基于观察来调制至少一个训练参数。 例如，本文描述的训练装置可以修改神经装置的内部学习机制（例如，尖峰速率，学习速率，神经调节器，传感器灵敏度等）和/或训练环境的刺激（例如，使火焰更靠近装置 ，使场景变暗等）。 以这种方式，与传统的神经装置训练系统相比，神经装置训练的速度（即，训练速率）可以显着增加。

公开(公告)号：WO2015148189A3

公开(公告)日：2015-12-17

申请号：PCT/US2015021077

申请日：2015-03-17

Applicant: QUALCOMM INC

Inventor： ANNAPUREDDY VENKATA SREEKANTA REDDY ,  JULIAN DAVID JONATHAN ,  TOWAL REGAN BLYTHE ,  LIU YINYIN

IPC: G06N3/04

CPC classification number: G06N3/0445 ,  G06N3/0481 ,  G06N3/049

Abstract: Differential encoding in a neural network includes predicting an activation value for a neuron in the neural network based on at least one previous activation value for the neuron. The encoding further includes encoding a value based on a difference between the predicted activation value and an actual activation value for the neuron in the neural network.

Abstract translation: 神经网络中的差分编码包括基于神经元的至少一个先前激活值来预测神经网络中神经元的激活值。 编码还包括基于预测的激活值和神经网络中的神经元的实际激活值之间的差来编码值。

公开(公告)号：WO2015112713A3

公开(公告)日：2015-10-29

申请号：PCT/US2015012454

申请日：2015-01-22

Applicant: QUALCOMM INC

Inventor： CANOY MICHAEL-DAVID NAKAYOSHI ,  LIU YINYIN ,  CHAN VICTOR HOKKIU ,  CAMPOS MICHAEL ,  LEVIN JEFFREY ALEXANDER ,  WIERZYNSKI CASIMIR MATTHEW

IPC: G06N3/04 ,  G06N3/063 ,  G06N99/00

CPC classification number: G06F9/52 ,  G06N3/049 ,  G06N3/063 ,  G06N3/08 ,  G06N99/005

Abstract: Values are synchronized across processing blocks in a neural network by encoding spikes in a first processing block with a value to be shared across the neural network. The spikes may be transmitted to a second processing block in the neural network via an interblock interface. The received spikes are decoded in the second processing block so as to generate a value that is synchronized with the value of the first processing block.

Abstract translation: 通过在第一处理块中对尖峰进行编码以跨神经网络共享的值来在神经网络中的处理块之间同步数值。 尖峰可以经由帧间接口发送到神经网络中的第二处理块。 接收的尖峰在第二处理块中解码，以产生与第一处理块的值同步的值。