公开(公告)号：US09436909B2

公开(公告)日：2016-09-06

申请号：US13922143

申请日：2013-06-19

Applicant: Brain Corporation

Inventor： Filip Piekniewski ,  Vadim Polonichko ,  Eugene Izhikevich

IPC: G06N5/00 ,  G06F1/00 ,  G06N3/08 ,  G06N3/04

CPC classification number: G06N3/08 ,  G06N3/049

Abstract: Apparatus and methods for processing inputs by one or more neurons of a network. The neuron(s) may generate spikes based on receipt of multiple inputs. Latency of spike generation may be determined based on an input magnitude. Inputs may be scaled using for example a non-linear concave transform. Scaling may increase neuron sensitivity to lower magnitude inputs, thereby improving latency encoding of small amplitude inputs. The transformation function may be configured compatible with existing non-scaling neuron processes and used as a plug-in to existing neuron models. Use of input scaling may allow for an improved network operation and reduce task simulation time.

Abstract translation: 用于由网络的一个或多个神经元处理输入的装置和方法。 基于多个输入的接收，神经元可以产生尖峰。 可以基于输入幅度来确定尖峰生成的延迟。 可以使用例如非线性凹变换来缩放输入。 缩放可以将神经元灵敏度增加到较低幅度的输入，从而改善小振幅输入的延迟编码。 转换函数可以被配置为与现有的非缩放神经元过程兼容，并且用作现有神经元模型的插件。 使用输入缩放可以允许改进的网络操作并减少任务模拟时间。

公开(公告)号：US20160075017A1

公开(公告)日：2016-03-17

申请号：US14489373

申请日：2014-09-17

Applicant: BRAIN CORPORATION

Inventor： Patryk Laurent ,  Csaba Petre ,  Eugene M. Izhikevich ,  Vadim Polonichko

IPC: B25J9/16 ,  B25J9/00 ,  B25J13/08

CPC classification number: B25J9/163 ,  G05B2219/2642 ,  G05B2219/40116 ,  Y10S901/05

Abstract: Computerized appliances may be operated by users remotely. In one implementation, a learning controller apparatus may be operated to determine association between a user indication and an action by the appliance. The user indications, e.g., gestures, posture changes, audio signals may trigger an event associated with the controller. The event may be linked to a plurality of instructions configured to communicate a command to the appliance. The learning apparatus may receive sensory input conveying information about robot's state and environment (context). The sensory input may be used to determine the user indications. During operation, upon determine the indication using sensory input, the controller may cause execution of the respective instructions in order to trigger action by the appliance. Device animation methodology may enable users to operate computerized appliances using gestures, voice commands, posture changes, and/or other customized control elements.

Abstract translation: 计算机化设备可由用户远程操作。 在一个实现中，可以操作学习控制器设备以确定用户指示与设备的动作之间的关联。 用户指示，例如手势，姿势改变，音频信号可以触发与控制器相关联的事件。 事件可以被链接到被配置为向设备传达命令的多个指令。 学习装置可以接收关于机器人的状态和环境（上下文）的感官输入。 感官输入可用于确定用户指示。 在操作期间，在使用感觉输入确定指示时，控制器可以引起相应指令的执行，以触发设备的动作。 设备动画方法可以使用户能够使用手势，语音命令，姿势改变和/或其他定制的控制元素来操作计算机化的设备。

公开(公告)号：US09208432B2

公开(公告)日：2015-12-08

申请号：US13830398

申请日：2013-03-14

Applicant: Brain Corporation

Inventor： Marius Buibas ,  Eugene M. Izhikevich ,  Botond Szatmary ,  Vadim Polonichko

IPC: G06N5/00 ,  G06F1/00 ,  G06N3/08 ,  G06N3/04 ,  G06N3/10

CPC classification number: G06N3/08 ,  G06N3/04 ,  G06N3/049 ,  G06N3/10

Abstract: Apparatus and methods for learning and training in neural network-based devices. In one implementation, the devices each comprise multiple spiking neurons, configured to process sensory input. In one approach, alternate heterosynaptic plasticity mechanisms are used to enhance learning and field diversity within the devices. The selection of alternate plasticity rules is based on recent post-synaptic activity of neighboring neurons. Apparatus and methods for simplifying training of the devices are also disclosed, including a computer-based application. A data representation of the neural network may be imaged and transferred to another computational environment, effectively copying the brain. Techniques and architectures for achieve this training, storing, and distributing these data representations are also disclosed.

Abstract translation: 基于神经网络的设备学习和训练的装置和方法。 在一个实施方式中，每个装置包括多个加标神经元，其配置成处理感觉输入。 在一种方法中，使用交替的异质突触可塑性机制来增强装置内的学习和场分集。 替代可塑性规则的选择是基于最近邻近神经元的突触后活动。 还公开了用于简化设备训练的装置和方法，包括基于计算机的应用。 神经网络的数据表示可以被成像并传送到另一个计算环境，有效地复制大脑。 还公开了用于实现这种训练，存储和分发这些数据表示的技术和架构。

公开(公告)号：US20150338204A1

公开(公告)日：2015-11-26

申请号：US14285414

申请日：2014-05-22

Applicant: Brain Corporation

Inventor： Micah Richert ,  Marius Buibas ,  Vadim Polonichko

IPC: G01B11/14 ,  H04N19/51

CPC classification number: G01B11/14 ,  G06T7/593 ,  G06T2207/10021 ,  G06T2207/30196 ,  G06T2207/30252 ,  H04N19/51

Abstract: Data streams from multiple image sensors may be combined in order to form, for example, an interleaved video stream, which can be used to determine distance to an object. The video stream may be encoded using a motion estimation encoder. Output of the video encoder may be processed (e.g., parsed) in order to extract motion information present in the encoded video. The motion information may be utilized in order to determine a depth of visual scene, such as by using binocular disparity between two or more images by an adaptive controller in order to detect one or more objects salient to a given task. In one variant, depth information is utilized during control and operation of mobile robotic devices.

Abstract translation: 可以组合来自多个图像传感器的数据流，以便形成例如交织的视频流，其可以用于确定到对象的距离。 可以使用运动估计编码器对视频流进行编码。 可以处理（例如，解析）视频编码器的输出，以便提取编码视频中存在的运动信息。 可以利用运动信息来确定视觉场景的深度，例如通过由自适应控制器使用两个或更多图像之间的双目视差来检测突出给定任务的一个或多个对象。 在一个变型中，在移动机器人设备的控制和操作期间利用深度信息。