公开(公告)号：US4718092A

公开(公告)日：1988-01-05

申请号：US593893

申请日：1984-03-27

申请人： John W. Klovstad

发明人： John W. Klovstad

IPC分类号： G10L15/18 ,  G10L5/00

CPC分类号： G10L15/193

摘要： Speech recognition calculations are decreased by deactivating (or activating) a word in a grammar graph at the "kernel" level. A word is a sequence of acoustic kernels, each kernel a phoneme spectral vector with min-max duration data on a template.

摘要翻译： 通过在“内核”级别中停用（或激活）语法图中的单词来减少语音识别计算。 一个单词是声学核的序列，每个核心是在模板上具有最小 - 最大持续时间数据的音素谱矢量。

公开(公告)号：US5603240A

公开(公告)日：1997-02-18

申请号：US206651

申请日：1994-03-04

申请人： John W. Klovstad ,  James A. Fortune

发明人： John W. Klovstad ,  James A. Fortune

IPC分类号： F16H29/20 ,  F16H53/06 ,  F16H29/16

CPC分类号： F16H53/06 ,  F16H29/20 ,  Y10T74/1515 ,  Y10T74/1516 ,  Y10T74/1523

摘要： The transmission couples a cam shaft to an output shaft through a plurality of sets of specialized linkages. Each set of specialized linkages comprises: a displacement waveform generator (DWG) comprising a solid cam rigidly attached to the cam shaft, the outer surface of which produces a periodic displacement that is linearly proportional to rotation angle of the cam shaft over a substantial part of a rotation cycle; an input linear displacement multiplier (ILDM) comprising a link surrounding the cam shaft such that when the cam shaft rotates the link oscillates back and forth with a displacement that is linearly proportional to rotation angle of the cam shaft over a substantial part of a rotation cycle; an output linear displacement multiplier (OLDM) comprising a link pivotally connected to the ILDM through a power coupler, able to slide back and forth, and having a longitudinally extending rack of gear teeth; a disk free to rotate coaxially with the output shaft that has a concentric pinion gear continuously engaging the rack of teeth along the OLDM such that the disk is rotationally oscillated as the OLDM moves back and forth; and one or more calipers that may lock the disk to a flange that is affixed to the transmission's output shaft. Disks are locked to flanges at such times that the output shaft is rotated in a uniform manner. The apparent gear ratio is continuously varied from full forward, to neutral, to reverse by changing the angle between the ILDM and OLDM.

摘要翻译： 传动装置通过多组专用联动装置将凸轮轴连接到输出轴。 每套专门的联动装置包括：位移波形发生器（DWG），其包括刚性地连接到凸轮轴的固体凸轮，其外表面产生与凸轮轴的旋转角线性比例的周期性位移， 旋转周期; 输入线性位移乘法器（ILDM），包括围绕凸轮轴的连杆，使得当凸轮轴旋转时，连杆以相对于旋转周期的大部分的凸轮轴的旋转角线性成比例的位移来回摆动 ; 输出线性位移乘法器（OLDM），其包括通过功率耦合器枢转地连接到ILDM的连杆，能够来回滑动并具有纵向延伸的齿轮齿; 与输出轴同轴旋转的盘，该输出轴具有沿OLDM连续接合齿条的同心小齿轮，使得当OLDM前后移动时，盘旋转振荡; 以及可以将盘锁定到固定到变速器的输出轴上的凸缘的一个或多个卡钳。 在这样的时间，将盘锁定到法兰上，使得输出轴以均匀的方式旋转。 通过改变ILDM和OLDM之间的角度，表观齿轮比从完全向前，中性到反向连续变化。

公开(公告)号：US4718088A

公开(公告)日：1988-01-05

申请号：US593891

申请日：1984-03-27

申请人： James K. Baker ,  John W. Klovstad ,  Chin-Hui Lee ,  Kalyan Ganesan

发明人： James K. Baker ,  John W. Klovstad ,  Chin-Hui Lee ,  Kalyan Ganesan

IPC分类号： G10L11/02 ,  G10L15/02 ,  G10L15/06 ,  G10L15/08 ,  G10L15/12 ,  G10L5/00

CPC分类号： G10L25/87 ,  G10L15/063 ,  G10L15/083 ,  G10L15/02 ,  G10L15/12 ,  G10L2015/0638 ,  G10L25/27

摘要： A speech recognition method and apparatus employ a speech processing circuitry for repetitively deriving from a speech input, at a frame repetition rate, a plurality of acoustic parameters. The acoustic parameters represent the speech input signal for a frame time. A plurality of template matching and cost processing circuitries are connected to a system bus, along with the speech processing circuitry, for determining, or identifying, the speech units in the input speech, by comparing the acoustic parameters with stored template patterns. The apparatus can be expanded by adding more template matching and cost processing circuitry to the bus thereby increasing the speech recognition capacity of the apparatus. Template pattern generation is advantageously aided by using a "joker" word to specify the time boundaries of utterances spoken in isolation, by finding the beginning and ending of an utterance surrounded by silence.

摘要翻译： 语音识别方法和装置采用语音处理电路，以帧重复率重复地从语音输入中导出多个声学参数。 声学参数表示帧时间的语音输入信号。 通过将声学参数与存储的模板图案进行比较，多个模板匹配和成本处理电路连同语音处理电路连接到用于确定或识别输入语音中的语音单元的系统总线。 可以通过向总线添加更多的模板匹配和成本处理电路来扩展该装置，从而增加装置的语音识别能力。 通过使用“小丑”字通过找到由沉默包围的话语的开始和结束来有助于指定孤立地说出的话语的时间边界。

公开(公告)号：US4713777A

公开(公告)日：1987-12-15

申请号：US593892

申请日：1984-05-27

申请人： John W. Klovstad ,  Chin-Hui Lee ,  Kalyan Ganesan

发明人： John W. Klovstad ,  Chin-Hui Lee ,  Kalyan Ganesan

IPC分类号： G10L15/00 ,  G10L5/00

CPC分类号： G10L15/00

摘要： In a speech recognition system, the beginning of speech versus non-speech (a cough or noise) is distinguished by reverting to a non-speech decision process whenever the liklihood cost of template (vocabulary) patterns, including silence, is worse than a predetermined threshold, established by a Joker Word which represents a non-vocabulary word score and path in the grammar graph.

摘要翻译： 在语音识别系统中，每当包括沉默在内的模板（词汇）模式的可能成本比预定的语言更差时，语音与非语音（咳嗽或噪音）的开始被归结为非语音决策过程 阈值，由Joker Word建立，表示语法图中的非词汇单词得分和路径。