返回项目页面
对比文件列表
2009-01-27_US7483868B_发明授权_US07483868B2 Automatic neural-net model generation and maintenance_+++c_i+++.docx
2026-03-19 23:20
2011-01-27_US2011022230A_发明申请_US20110022230A1 HYBRID CONTROL DEVICE_+++A_C_F_I_b_d_e_g_h+++.docx
2026-03-19 23:20
2012-02-28_US8126828B_发明授权_US08126828B2 Special purpose processor implementing a synthetic neural model of the human brain_+++C_F_I_a_b_g+++.docx
2026-03-19 23:20
2012-05-03_US2012109863A_发明申请_US20120109863A1 CANONICAL SPIKING NEURON NETWORK FOR SPATIOTEMPORAL ASSOCIATIVE MEMORY_+++C_D_I+++.docx
2026-03-19 23:20
2012-06-27_CN102521653A_发明公开_CN102521653A 井下多机器人联合搜救的生物刺激神经网络设备及其方法_+++A_B_C_F+++.docx
2026-03-19 23:20
2012-10-04_WO2012130251A_发明申请_WO2012130251A1 IMAGE UNDERSTANDING BASED ON FUZZY PULSE - COUPLED NEURAL NETWORKS_+++F_c_i+++.docx
2026-03-19 23:20
2012-12-11_US8332070B_发明授权_US08332070B2 Learning and use of schemata in robotic devices_+++C_F_I_a+++.docx
2026-03-19 23:20
2013-01-29_US8364311B_发明授权_US08364311B2 Driver assistance system or robot with dynamic attention module_+++A_F_c_h+++.docx
2026-03-19 23:20
2013-01-30_CN102906767A_发明公开_CN102906767A 用于时空联合存储器的标准尖峰神经网络_+++F_c_d_i+++.docx
2026-03-19 23:20
2013-03-21_US2013073501A_发明申请_US20130073501A1 METHOD AND APPARATUS FOR STRUCTURAL DELAY PLASTICITY IN SPIKING NEURAL NETWORKS_+++C_I+++.docx
2026-03-19 23:20
2013-08-08_US2013204820A_发明申请_US20130204820A1 METHODS AND APPARATUS FOR SPIKING NEURAL COMPUTATION_+++c_i+++.docx
2026-03-19 23:20
2013-11-12_US8583286B_发明授权_US08583286B2 Hybrid control device_+++D_F_I_a_b_c_e_g+++.docx
2026-03-19 23:20
US2005047647A1_Description_20260309_2113_+++C_F_G_d_i+++.docx
2026-03-19 23:20
US2005261803A1_Description_20260309_2113_+++A_C_F_I_b+++.docx
2026-03-19 23:20
2013-11-12_US8583286B_发明授权_US08583286B2 Hybrid control device_+++D_F_I_a_b_c_e_g+++.docx

对比文件名称:2013-11-12_US8583286B_发明授权_US08583286B2 Hybrid control device

目标专利名称:用于空间目标选择的失衡式交叉抑制性机制 CN106030621B

本次调用模型名称:专利创造性评估模型

### 特征比对表格

技术特征描述及公开性判断结果对比文件原文引用公开性论述
**技术特征A**:包括:至少部分地基于目标选择准则从所述多个目标中为机器人设备选择目标,<br>**《隐含公开》**说明书第[0070]段:“...the action selection module decides that the ball should be chased.” <br>说明书第[0078]段:“...a non-neural algorithm can be responsible for deciding what object is salient at any given time, and whether that object should be approached or merely tracked...” <br>说明书第[0091]段:“The amount of activity in a neuronal group can cause a non-neural controller to make a decision. For example, if summed activity of neuronal units in the Ball area exceeds a threshold the ball is present in the field of view, and the action selection unit may decide to chase the ball.”对比文件公开了机器人设备(SS-BBD)具有决策模块(动作选择模块),用于从多个对象(如足球、队友、对手、球门)中决定在任意给定时间哪个对象是显著的(salient),并决定是接近还是仅仅追踪该对象(第[0078]段)。决策的依据(即选择准则)可以是神经活动量超过阈值(第[0091]段)。因此,对比文件隐含公开了机器人设备基于某种选择准则(例如神经活动量阈值)从多个目标中选择一个目标(如足球)的过程。虽然对比文件的选择准则细节(如“显著”的判断)可能与目标专利不完全相同,但“基于准则选择目标”这一上位概念已被公开。
**技术特征B**:所述多个目标中的每个目标对应于包括相对于所述机器人设备的位置的多个位置的目标地图中的不同位置,<br>**《隐含公开》**说明书第[0042]段:“Darwin++ is a hybrid model that combines probabilistic localization algorithms with the ability of the brain region called the hippocampus to encode information about location.” <br>说明书第[0042]段:“The location of the device can be determined through a particle filter, which maintains many localization hypotheses (particles).” <br>说明书第[0044]段:“...the non-neural localization algorithms can produce an estimate of location on a map of the room that is encoded topographically in a neural area feeding the hippocampus.” <br>说明书第[0070]段:“The ball neuronal area can also recognize the ball by its shape: a conjunction of edges in a particular configuration. The object areas can have recurrent self-excitatory connections and some also had inhibitory connections to the other object areas.”对比文件公开了机器人设备(Darwin++)具有定位和地图构建能力(SLAM),能够估计自身在房间地图上的位置(第[0042], [0044]段)。同时,视觉系统包含多个对象识别神经区域(如Ball, Goal, Teammate, Opponent),这些区域的活动对应于视野中特定对象的存在和位置(第[0070]段)。这些对象区域接收来自视觉输入的拓扑映射(retinotopically),其活动模式实质上构成了一个以机器人自身为中心的“目标地图”,其中不同对象(目标)对应于该地图(视野)中的不同位置。虽然对比文件未明确使用“目标地图”一词,但本领域技术人员可以理解,基于视觉的、以自我为中心的、编码了多个目标位置的神经表示,构成了隐含公开的“目标地图”。
**技术特征C**:并且每个位置对应于人工神经网络中的神经元<br>**《隐含公开》**说明书第[0022]段:“The neuronal controller portion 112 can have a number of neuronal units. The neuronal units can emulate neurons or groups of neurons.” <br>说明书第[0070]段:“...object recognition can be based on color information, tuned self-excitation, and cross inhibition between the object groups...” <br>说明书表3:“Object Areas (5): Ball, OurGoal, TheirGoal, Teammate, Opponent (Size: 20 x 20)”对比文件详细描述了其人工神经网络包含多个神经区域和神经元单元(第[0022]段)。具体地,存在专门的对象区域(如Ball, OurGoal等),每个区域由多个神经元单元(20x20)组成(表3)。这些对象区域的活动对应于特定目标在视野中的位置(第[0070]段)。因此,每个目标的位置信息是通过对应对象神经区域中特定神经元单元的活动模式来编码的。这隐含公开了“每个位置对应于人工神经网络中的神经元”这一特征,即位置信息由神经元的激活模式来表征。
**技术特征D**:通过经由从对应于所选目标的目标神经元至不对应于所选目标的非目标神经元的连接输出抑制性权重来设置失衡<br>**《直接公开》**说明书第[0070]段:“The object areas can have recurrent self-excitatory connections and some also had inhibitory connections to the other object areas. Inhibitory connections make the system more robust: Object groups that should not be in the same place in the visual field can have inhibitory connections to each other.” <br>说明书第[0094]段:“A non-neural controller can modulate neural activity. In one version of the SS-BBD, a model was implemented where instead of having a temporary connection from the Ball group made to the Head group during ball-chasing, there was connections from all object areas to the Head group. However, the action selection module would increase the basal activity of the area that was of interest, and allow cross-inhibition between object areas to cancel out other signals to the Head motor area.” <br>说明书表4:“Ball→Teammate: cij(0) = -5.5, -5.6; Ball→Opponent: cij(0) = -5.5, -5.6; OurGoal→Teammate: cij(0) = -5.5, -5.6; ...” (负值表示抑制性连接)对比文件明确公开了在不同对象神经区域(如Ball, Teammate, Opponent等)之间存在抑制性连接(cross-inhibition)(第[0070]段,表4)。这些抑制性连接被描述为使系统更鲁棒,并用于取消(cancel out)向运动区域传递的其他信号(第[0094]段)。当动作选择模块决定关注某个对象(如足球)时,它会增加该对象区域的基础活动,并允许对象区域之间的交叉抑制来压制其他对象区域的信号。这直接对应于“通过从对应于所选目标的目标神经元(如Ball区域)至不对应于所选目标的非目标神经元(如Teammate、Opponent区域)的连接输出抑制性权重来设置失衡”。对比文件中的抑制性连接是实现对象选择/区分的关键机制,与目标专利中通过抑制性权重设置失衡以实现目标选择的作用相同。
**技术特征E**:以及至少部分地基于所述失衡来修改所述目标神经元与所述非目标神经元之间的相对激活,以使得在所述目标神经元的活跃量大于所述非目标神经元的活跃量时所述机器人设备移动至所述目标地图中与所选目标相关联的位置。<br>**《隐含公开》**说明书第[0075]段:“...activity in the pan area can drive the panning actuator in a topological fashion: more activity to the right of the area means the camera turns to the right. Thus the camera can turn to the right to re-center an object that is in the right half of the image. The tilt area can work similarly in the vertical direction.” <br>说明书第[0076]段:“The activity of the pan area can, in turn, project topographically to the body motor area, which controls the rotational velocity of the SS-BBD. Off center activity in the body area can create a turn in that direction. The total system can result in a cascade of activity that comes from visual areas to the motor areas, which in turn creates a change in behavior and changes in the visual input.” <br>说明书第[0091]段:“The amount of activity in a neuronal group can cause a non-neural controller to make a decision... the action selection unit may decide to chase the ball.” <br>结合特征D的论述,交叉抑制修改了相对激活。对比文件公开了对象神经区域的活动(经过交叉抑制调制后)被映射到控制摄像头转向和身体转动的运动神经区域(第[0075], [0076]段)。特定对象区域(如被选中的Ball区域)的活跃度更高,会导致运动区域产生相应的活动,进而驱动机器人设备(通过轮子)转向并接近该对象(第[0076]段,行为“Track Object”)。虽然对比文件未将“修改相对激活”与“移动至目标位置”用一句话直接关联,但本领域技术人员可以从其整体系统描述中毫无疑义地推断出:通过交叉抑制(特征D)使得被选中的目标对象神经区域相对于其他对象区域具有更高的活跃度,这种活跃度差异被传递到运动控制系统,最终导致机器人移动至与该选中目标相关联的位置(在视野中重新居中并接近它)。因此,该技术特征被隐含公开。
**技术特征F**:其特征在于,每个目标是空间目标。<br>**《直接公开》**说明书第[0067]段:“In one embodiment, there are five important objects on a robot sensor field with different colors or color combinations: our goal, opponent's goal, ball, teammate and opponent.” <br>说明书第[0070]段:“These object groups can receive their topographical inputs from one or more color neuronal groups.”对比文件明确列举了机器人传感器场中的重要对象:己方球门、对方球门、足球、队友和对手(第[0067]段)。这些对象都是存在于物理空间中的实体,其位置通过视觉系统进行拓扑编码(第[0070]段)。因此,每个目标都是空间目标,被对比文件直接公开。
**技术特征G**:其特征在于,所述目标选择准则至少部分地基于每个目标的活跃量概率。<br>**《隐含公开》**说明书第[0067]段:“The UV components (color information) of the image can be compared with previously generated lookup tables of salient colors. The lookup tables can represent the probability that a specific color in the image matches one of the salient colors.” <br>说明书第[0091]段:“The amount of activity in a neuronal group can cause a non-neural controller to make a decision. For example, if summed activity of neuronal units in the Ball area exceeds a threshold the ball is present in the field of view, and the action selection unit may decide to chase the ball.”对比文件公开了颜色查找表提供了图像中特定颜色与显著颜色匹配的“概率”(第[0067]段)。此外,决策(选择)可以基于神经区域的“活动量”(amount of activity)是否超过阈值(第[0091]段)。神经活动量本身可以视为目标存在或显著性的一个概率性或置信度指标。本领域技术人员可以合理推断,在选择目标时,可以(至少部分地)依据这种基于神经活动量或颜色匹配概率的准则。因此,该特征被隐含公开。
**技术特征H**:其特征在于,所述选择目标选择准则至少部分地基于所述目标与所述机器人设备之间的空间距离。<br>**未公开**说明书第[0081]-[0083]段描述了一个障碍物避免模型,该模型考虑了到障碍物的距离(`do`)和到目标物体的距离(`dt`)来计算角速度。然而,该模型用于路径规划(平衡接近目标和避开障碍物),而非用于从多个目标中选择哪一个作为要追踪或接近的目标。关于目标选择,对比文件未明确记载选择准则包括目标与机器人之间的距离。对比文件中公开的障碍物避免算法确实考虑了到目标的距离(`dt`)(第[0081]段),但这是在进行“Track Object”行为时,为了规划路径、避开障碍物而使用的信息。在更高层的“选择目标”决策(例如,决定是追踪球还是追踪对手)中,对比文件描述的依据是对象的显著性、神经活动量是否超过阈值、或者非神经算法基于传感器信息的决策(第[0078], [0091]段),并未提及将目标与机器人之间的空间距离作为选择准则的一部分。因此,对比文件未公开该特征。
**技术特征I**:其特征在于,所述连接是第一输入层连接、神经元输入、侧向连接、或其组合中的至少一者。<br>**《直接公开》**说明书第[0024]段描述了两种连接类型:电压无关(VI)和电压相关(VD)。<br>说明书表2和表4详细列出了大量投影(连接),包括从输入区域(如V1-color)到处理区域(如V2/4-color)的连接(可视为输入层连接),区域内神经元之间的连接(如`IT->IT`,可视为侧向连接),以及从处理区域到其他区域的连接(可视为神经元输入/输出)。例如表2中的 `V1-color→V2/4-color`(输入层连接),`IT→IT`(侧向连接),`ECIN→DG`(神经元输入)。对比文件详细描述了其人工神经网络中存在的各种连接类型。从输入传感器映射到初级神经区域的连接(如视觉输入到V1)对应于“第一输入层连接”;神经区域内部神经元之间的连接(如表2中的`IT→IT`)对应于“侧向连接”;以及一个神经区域的输出作为另一个神经区域的输入,这对应于“神经元输入”。对比文件公开的连接涵盖了这些类型,因此该特征被直接公开。

<<<D>>><<<F>>><<<I>>><<<a>>><<<b>>><<<c>>><<<e>>><<<g>>>

Powered by Django

网站备案号:渝ICP备2023012882号


重庆市非显而易见网络科技有限责任公司 A Anti NPE NPE