公开(公告)号：US07739110B2

公开(公告)日：2010-06-15

申请号：US11565628

申请日：2006-12-01

申请人： Chung-Hsien Wu ,  Yu-Sheng Lai ,  Chien-Lin Huang ,  Chia-Hao Kang

发明人： Chung-Hsien Wu ,  Yu-Sheng Lai ,  Chien-Lin Huang ,  Chia-Hao Kang

IPC分类号： G10L15/00 ,  G10L15/08 ,  G06F17/30

CPC分类号： G10L21/06 ,  G10L15/26

摘要： A method and an apparatus for multimedia data management are disclosed. The method provides an indexing and retrieval scheme for digital photos with speech annotations based on image-like patterns transformed from the recognized syllable candidates. For annotated spoken content, the recognized n-best syllable candidates are transformed into a sequence of syllable-transformed patterns. Eigen-image analysis is further adopted to extract the significant information to reduce the dimensionality. Vector quantization is applied to quantize the syllable-transformed patterns into feature vectors for indexing. The invention of indexing scheme reduces the dimensionality and noise of data, and achieves better performance of 16.26% for speech annotated photo retrieval.

摘要翻译： 公开了一种用于多媒体数据管理的方法和装置。 该方法为基于从识别的音节候选者转换的图像样式的语音注释的数字照片提供索引和检索方案。 对于有注释的口头内容，识别的最好的音节候选人被转换成音节转换模式的序列。 进一步采用特征图像分析提取重要信息以降低维数。 矢量量化被应用于将音节变换模式量化为用于索引的特征向量。 索引方案的发明降低了数据的维度和噪声，并且对于语音注释的照片检索实现了更高的16.26％的性能。

公开(公告)号：US20080256376A1

公开(公告)日：2008-10-16

申请号：US11961487

申请日：2007-12-20

申请人： Yi-Ping You ,  Jeng Kuen Lee ,  Kuo Yu Chuang ,  Chung-Hsien Wu

发明人： Yi-Ping You ,  Jeng Kuen Lee ,  Kuo Yu Chuang ,  Chung-Hsien Wu

IPC分类号： G06F1/32

CPC分类号： G06F1/3203 ,  G06F1/3246 ,  G06F9/30072 ,  G06F9/30083 ,  G06F9/3851 ,  Y02D50/20

摘要： The invention relates to a multi-thread power gating control design, setting idle components into a sleep mode to reduce power consumption due to current leakage. Based on compiler techniques, the invention arranges predicted-power-gating instructions into every thread of a may-happen-in-parallel region. A predicted-power-on instruction determines whether the corresponding component has been powered on, and powers on the component when it has not been powered on yet. A predicted-power-off instruction determines whether the component is required in the rest of the may-happen-in-parallel region, and powers off the component when it is required later.

摘要翻译： 本发明涉及一种多线程门控控制设计，将空闲组件设置为睡眠模式，以减少由于电流泄漏引起的功耗。 基于编译器技术，本发明将预测功率门控指令布置在可能发生并行区域的每个线程中。 预测开机指令确定相应的组件是否已经通电，并且当组件尚未通电时，对组件通电。 预测电源关闭指令确定组件是否在可能发生并行区域的其余部分中是必需的，并且在稍后需要时关闭组件。

公开(公告)号：US06380587B1

公开(公告)日：2002-04-30

申请号：US09706527

申请日：2000-11-03

申请人： Shing-Ren Sheu ,  Chung-Hsien Wu ,  Chih-Ming Huang

发明人： Shing-Ren Sheu ,  Chung-Hsien Wu ,  Chih-Ming Huang

IPC分类号： H01L2976

CPC分类号： H01L27/1126 ,  H01L27/105 ,  H01L27/11293

摘要： A semiconductor read-only memory (ROM) and a method of fabricating the same are provided. The ROM device is structured in such a manner that allows the fabrication to include a fewer number of mask processes. This makes it more cost effective and allows a cycle time that is shorter than that of the prior art. Moreover, the particular structure of the ROM device makes punchthrough less likely to occur between any neighboring pairs of the buried bit lines when the ROM device is further scaled down. The ROM device is constructed on a semiconductor substrate which is partitioned into a peripheral region and a cell region. A plurality of STI structures are formed at predefined locations in both the peripheral region and the cell region. Immediately after this, a first ion-implantation process can be performed on the cell region to form a plurality of buried bit lines. Subsequently, the dielectric isolation layers in all of the STI structures in the cell region are removed, leaving a plurality of empty trenches behind. A conformal insulating layer and a conductive layer are then successively formed over the wafer, and the conductive layer is further selectively removed to form a word line in the cell region and a gate in the peripheral region. In the code implantation process, selected channel regions between the buried bit lines are doped with impurities for code implantation of data into the ROM device.

公开(公告)号：US6103608A

公开(公告)日：2000-08-15

申请号：US62661

申请日：1998-04-20

申请人： Yi-Min Jen ,  Chung-Hsien Wu

发明人： Yi-Min Jen ,  Chung-Hsien Wu

IPC分类号： H01L21/60 ,  H01L21/336 ,  H01L21/4763

CPC分类号： H01L21/76897

摘要： The present invention discloses a method of forming a contact window on a substrate. The method in the present invention includes a step of forming a gate structure on said substrate having a gate oxide, a gate electrode on the gate oxide, and a gate electrode protection layer on the gate electrode, a step of forming a protection layer conforming with the substrate and the gate structure, and a step of forming a first insulation layer over the protection layer. The method further includes removing a portion of the first insulation layer and a portion of the protection layer for forming side wall spacers at side walls of the gate structure, performing a ion implantation to the substrate using the gate structure and the side wall spacers as a mask, and then forming a second insulation layer on the substrate, the side wall spacers, and the gate structure. The method further includes patterning a photoresist layer on the second insulation layer to expose a portion of second insulation layer for making a connection. Finally, a portion of the second insulation layer is removed for forming a contact window.

摘要翻译： 本发明公开了一种在基板上形成接触窗的方法。 本发明的方法包括：在具有栅极氧化物的栅极结构上形成栅极结构的步骤，栅极氧化物上的栅电极和栅电极上的栅电极保护层，形成与 基板和栅极结构，以及在保护层上形成第一绝缘层的步骤。 该方法还包括去除第一绝缘层的一部分和用于在栅极结构的侧壁处形成侧壁间隔物的保护层的一部分，使用栅极结构和侧壁间隔物作为离子注入到基底 掩模，然后在衬底，侧壁间隔物和栅极结构上形成第二绝缘层。 该方法还包括在第二绝缘层上图案化光致抗蚀剂层，以露出用于进行连接的第二绝缘层的一部分。 最后，去除第二绝缘层的一部分以形成接触窗。

公开(公告)号：US08478591B2

公开(公告)日：2013-07-02

申请号：US12637990

申请日：2009-12-15

申请人： Huan-Chung Li ,  Chung-Hsien Wu ,  Han-Ping Shen ,  Chun-Kai Wang ,  Chia-Hsin Hsieh

发明人： Huan-Chung Li ,  Chung-Hsien Wu ,  Han-Ping Shen ,  Chun-Kai Wang ,  Chia-Hsin Hsieh

IPC分类号： G10L15/06 ,  G10L15/00

CPC分类号： G10L15/065 ,  G10L15/187

摘要： A phonetic variation model building apparatus, having a phoneme database for recording at least a standard phonetic model of a language and a plurality of non-standardized phonemes of the language is provided. A phonetic variation identifier identifies a plurality of phonetic variations between the non-standardized phonemes and the standard phonetic model. A phonetic transformation calculator calculates a plurality of coefficients of a phonetic transformation function based on the phonetic variations and the phonetic transformation function. A phonetic variation model generator generates at least a phonetic variation model based on the standard phonetic model, the phonetic transformation function and the coefficients thereof.

摘要翻译： 提供了一种具有用于记录语言的至少标准语音模型和语言的多个非标准化音素的音素数据库的语音变体模型构建装置。 语音变化标识符识别非标准化音素和标准语音模型之间的多个语音变化。 语音变换计算器基于语音变化和语音转换函数来计算语音转换函数的多个系数。 语音变异模型发生器基于标准语音模型，语音变换函数及其系数产生至少一个语音变化模型。

公开(公告)号：US08057781B2

公开(公告)日：2011-11-15

申请号：US12813403

申请日：2010-06-10

申请人： Chung-Hsin Lu ,  Chung-Hsien Wu ,  Szu-Chia Chien ,  Zhi-Liang Liu

发明人： Chung-Hsin Lu ,  Chung-Hsien Wu ,  Szu-Chia Chien ,  Zhi-Liang Liu

IPC分类号： C01B17/00 ,  C01B19/00

CPC分类号： C01B19/007 ,  C01B19/002 ,  C01P2002/72 ,  C01P2004/61 ,  C01P2004/62 ,  Y10S977/814 ,  Y10S977/815

摘要： The invention provides a fabrication method for a chalcopyrite powder. The fabrication method includes: (a) mixing a Group IB compound and a Group IIIA compound in a solvent; (b) drying or precipitating the solution of step (a) to obtain a precursor containing Group IB and Group IIIA elements; (c) mixing a solution or powder containing a Group VIA compound with the precursor; and (d) heating the mixture of step (c) to obtain the chalcopyrite powder.

摘要翻译： 本发明提供了一种黄铜矿粉末的制造方法。 制造方法包括：（a）在溶剂中混合IB族化合物和IIIA族化合物; （b）干燥或沉淀步骤（a）的溶液以获得含有IB和IIIA族元素的前体; （c）将含有VIA族化合物的溶液或粉末与前体混合; 和（d）加热步骤（c）的混合物以获得黄铜矿粉末。

公开(公告)号：US08000964B2

公开(公告)日：2011-08-16

申请号：US12034842

申请日：2008-02-21

申请人： Jui-Tang Lin ,  Chin-Shun Hsu ,  Shen-Yen Chai ,  Chung-Hsien Wu ,  Kuei-Ming Lee ,  Chia-Hsin Hsieh ,  Chien-Lin Huang

发明人： Jui-Tang Lin ,  Chin-Shun Hsu ,  Shen-Yen Chai ,  Chung-Hsien Wu ,  Kuei-Ming Lee ,  Chia-Hsin Hsieh ,  Chien-Lin Huang

IPC分类号： G10L15/04

CPC分类号： G10L15/187

摘要： A method of constructing a model of recognizing English pronunciation variations is used to recognize English pronunciations with different intonations influenced by non-English native languages. The method includes collecting a plurality of sound information corresponding to English expressions; corresponding phonetic alphabets of the non-English native language and English of a region to International Phonetic Alphabets (IPAs), so as to form a plurality of pronunciation models; converting the sound information with the pronunciation models to form a pronunciation variation network of the corresponding English expressions, thereby detecting whether the English expressions have pronunciation variation paths; and finally summarizing the pronunciation variation paths to form a plurality of pronunciation variation rules. Furthermore, the pronunciation variations are represented by phonetics features to infer possible pronunciation variation rules, which are stored to form pronunciation variation models. The construction of the pronunciation variation models enhances applicability of an English recognition system and accuracy of voice recognition.

摘要翻译： 构建识别英文发音变体的模型的方法用于识别受非英语母语影响的不同语调的英文发音。 该方法包括收集对应于英文表达的多个声音信息; 相应的非英语母语的语音字母和一个区域的英语到国际语音字母（IPAs），以形成多个发音模型; 用声音模型转换声音信息，形成相应英文表达的发音变体网络，从而检测英文表达是否具有发音变化路径; 并且最后总结发音变化路径以形成多个发音变化规则。 此外，发音变化由声音特征表示，以推断可能的发音变化规则，其被存储以形成发音变体模型。 发音变异模型的构建增强了英语识别系统的适用性和语音识别的准确性。

公开(公告)号：US20070288237A1

公开(公告)日：2007-12-13

申请号：US11565628

申请日：2006-12-01

申请人： Chung-Hsien Wu ,  Yu-Sheng Lai ,  Chien-Lin Huang ,  Chia-Hao Kang

发明人： Chung-Hsien Wu ,  Yu-Sheng Lai ,  Chien-Lin Huang ,  Chia-Hao Kang

IPC分类号： G10L15/26

CPC分类号： G10L21/06 ,  G10L15/26

摘要： A method and an apparatus for multimedia data management are disclosed. The method provides an indexing and retrieval scheme for digital photos with speech annotations based on image-like patterns transformed from the recognized syllable candidates. For annotated spoken content, the recognized n-best syllable candidates are transformed into a sequence of syllable-transformed patterns. Eigen-image analysis is further adopted to extract the significant information to reduce the dimensionality. Vector quantization is applied to quantize the syllable-transformed patterns into feature vectors for indexing. The invention of indexing scheme reduces the dimensionality and noise of data, and achieves better performance of 16.26% for speech annotated photo retrieval.

摘要翻译： 公开了一种用于多媒体数据管理的方法和装置。 该方法为基于从识别的音节候选者转换的图像样式的语音注释的数字照片提供索引和检索方案。 对于有注释的口头内容，识别的最好的音节候选人被转换成音节转换模式的序列。 进一步采用特征图像分析提取重要信息以降低维数。 矢量量化被应用于将音节变换模式量化为用于索引的特征向量。 索引方案的发明降低了数据的维度和噪声，并且对于语音注释的照片检索实现了更高的16.26％的性能。

公开(公告)号：US07292725B2

公开(公告)日：2007-11-06

申请号：US11069327

申请日：2005-03-01

申请人： Chih-Lung Chen ,  Chien-Hong Lin ,  Chung-Hsien Wu ,  Zong-Yau Yin ,  Chi-Lung Wang

发明人： Chih-Lung Chen ,  Chien-Hong Lin ,  Chung-Hsien Wu ,  Zong-Yau Yin ,  Chi-Lung Wang

IPC分类号： G06K9/00

CPC分类号： H04N9/045 ,  G06T3/4015 ,  H04N9/07

摘要： A demosaicking method and apparatus for color filter array interpolation in digital image acquisition systems is provided. The method includes a data-training phase and a data-practice phase. In the data-training phase of this invention, a huge volume of image samples are received for statistical analysis and compared with their corresponding full-color images. The optimal samples are then recorded in a database. In the data-practice phase of this invention, the images of color filter array are reconstructed. The present invention not only significantly improves the image quality but also reduces undue hardware cost. Furthermore, it has an advantage of ease of hardware implementation. All the color relationships of neighboring pixels are recorded in a database for statistical analysis. The database is not limited to only one set. It is also pretty flexible, and can be adjusted or updated depending on desired applications.

摘要翻译： 提供了一种用于数字图像采集系统中滤色器阵列插值的去马赛克方法和装置。 该方法包括数据训练阶段和数据实践阶段。 在本发明的数据训练阶段，接收大量图像样本用于统计分析，并与其相应的全色图像进行比较。 然后将最佳样本记录在数据库中。 在本发明的数据实践阶段，重构滤色器阵列的图像。 本发明不仅显着地提高了图像质量，而且降低了不必要的硬件成本。 此外，它具有易于实现硬件的优点。 相邻像素的所有颜色关系记录在数据库中用于统计分析。 数据库不仅限于一个集合。 它也非常灵活，可以根据所需的应用进行调整或更新。