专利快速检索-快速检索全球专利，免费商用专利数据库-IPRDB

1. 发明申请

WO2014207752A1 AUTHENTICATING USER BY CORRELATING SPEECH AND CORRESPONDING LIP SHAPE 审中-公开
标题翻译：通过相关语音和相关的LIP形状来认证用户
公开(公告)号：WO2014207752A1
公开(公告)日：2014-12-31
申请号：PCT/IN2013/000397
申请日：2013-06-27
申请人： HEWLETT-PACKARD DEVELOPMENT COMPANY, L.P. , RAMACHANDRULA, Sitaram , RAVISHANKAR, Hariharan
发明人： RAMACHANDRULA, Sitaram , RAVISHANKAR, Hariharan
IPC分类号： G06K9/00
CPC分类号： G06K9/6293 , G06K9/00335 , G06K9/00892 , G06K9/629 , G06T7/73 , G06T2207/10016 , G06T2207/30201 , G10L17/10 , G10L19/038 , G10L25/51 , G10L2019/0005
摘要： Provided is a method of authenticating a user by correlating speech and corresponding lip shape. An audiovisual of a user requesting authentication is captured. The audiovisual is processed to generate a speech vector quantization sequence and a corresponding lip vector quantization sequence of the user. A likelihood of the speech vector quantization sequence and the corresponding lip vector quantization sequence with probability distributions of speech vector quantization code words corresponding to different lip shape vector quantization code words of the user requesting authentication weighed by probabilities of speech and lip vector quantization indices of the user requesting authentication is evaluated. If upon evaluation, a likelihood of the user requesting authentication being an authentic user is more than a predefined threshold, the user is authenticated.
摘要翻译：提供了通过将语音和对应的唇形状相关联来认证用户的方法。捕获请求认证的用户的视听。视听被处理以产生用户的语音矢量量化序列和相应的唇矢量量化序列。语音矢量量化序列的可能性和对应于用户请求认证的不同唇形矢量量化码字的语音矢量量化码字的概率分布的相应唇矢量量化序列被语音和嘴唇矢量量化索引的概率加权评估用户请求认证。如果在评估时，用户请求认证是可信用户的可能性大于预定阈值，则认证用户。

2. 发明申请

WO2007017883A1 SYSTEM AND METHOD FOR VOICE ASSISTED INPUTTING OF SYLLABIC CHARACTERS INTO A COMPUTER 审中-公开
标题翻译：用于语音辅助将字符串输入计算机的系统和方法
公开(公告)号：WO2007017883A1
公开(公告)日：2007-02-15
申请号：PCT/IN2005/000258
申请日：2005-08-05
申请人： HEWLETT-PACKARD DEVELOPMENT COMPANY L.P. , RAMACHANDRULA, Sitaram , BALI, Kalika
发明人： RAMACHANDRULA, Sitaram , BALI, Kalika
IPC分类号： G06F3/023 , G06F3/00
CPC分类号： G06F3/0233 , G06F3/018 , G06F2203/0381
摘要： A technique to input characters of syllabic scripts in a practical and natural way into a computer that includes substantially simultaneously inputting a base character along with voice data of a desired syllable associated with the base character via a voice producing device such as a microphone. In one example embodiment, this is achieved by locating a base character associated with a desired syllable from a set of characters disposed in a character input device. An associated data of the located base character is then inputted using the character input device into a voice-recognition engine. The voice data associated with the spoken desired syllable is then inputted substantially simultaneously into the voice-recognition engine. A current candidate syllable is then produced by the voice-recognition engine as a function of the inputted associated data and the substantially simultaneously inputted voice data.
摘要翻译：一种以实用和自然的方式将音节脚本的字符输入到计算机中的技术，该计算机包括基本上同时输入基本字符以及通过诸如麦克风的语音产生设备与基本字符相关联的期望音节的语音数据。在一个示例实施例中，这是通过从设置在字符输入设备中的一组字符定位与期望音节相关联的基本字符来实现的。然后，使用字符输入装置将定位的基本字符的关联数据输入到语音识别引擎中。然后，与口语期望音节相关联的语音数据基本上同时输入到语音识别引擎中。语音识别引擎随后根据输入的相关数据和基本上同时输入的语音数据产生当前候选音节。

3. 发明申请

WO2011001439A2 SKEW DETECTION 审中-公开
标题翻译： SKEW检测
公开(公告)号：WO2011001439A2
公开(公告)日：2011-01-06
申请号：PCT/IN2009/000373
申请日：2009-07-02
申请人： HEWLETT-PACKARD DEVELOPMENT COMPANY, L.P. , RAMACHANDRULA, Sitaram , BANERJEE, Serene , PARIKH, Pulkit , SHAFFI, Noushath , KUCHIBHOTLA, Anjaneyulu, Seetha, Rama , DEY, Prasenjit
发明人： RAMACHANDRULA, Sitaram , BANERJEE, Serene , PARIKH, Pulkit , SHAFFI, Noushath , KUCHIBHOTLA, Anjaneyulu, Seetha, Rama , DEY, Prasenjit
IPC分类号： G06T7/40
CPC分类号： H04N1/3878 , G06K9/3283 , G06K9/4633 , G06K9/4647 , G06K2209/01 , G06T7/12 , G06T7/194 , G06T7/70 , G06T2207/20061 , G06T2207/30176
摘要： Presented is a skew detection apparatus. In one form, the apparatus estimates skew based on the locations of a set of foreground content pixels or a set of edge pixels that are nearest to the side of an image of a document. In another form, the apparatus includes a skew estimation unit adapted to estimate skew based on the orientation of foreground or background content in the interior of a document. In another form, the apparatus estimates skew using segments of an image of a document. Also presented is a document image processing apparatus including the skew detection apparatus.
摘要翻译：提出了一种偏斜检测装置。在一种形式中，该装置基于一组前景内容像素或最靠近文档图像侧边缘像素的集合的位置来估计偏差。在另一种形式中，该装置包括一个歪斜估计单元，用于基于文档内部的前景或背景内容的方向来估计偏差。在另一种形式中，该装置使用文档的图像的段来估计偏斜。还提出了包括歪斜检测装置的文件图像处理装置。

4. 发明申请

WO2007000773A1 METHOD AND SYSTEM FOR NAVIGATING AND EDITING TELEPHONIC INTERACTIVE VOICE RESPONSE INFORMATION 审中-公开
标题翻译：用于导航和编辑电话交互式语音响应信息的方法和系统
公开(公告)号：WO2007000773A1
公开(公告)日：2007-01-04
申请号：PCT/IN2005/000218
申请日：2005-06-28
申请人： HEWLETT - PACKARD DEVELOPMENT COMPANY, L.P. , RAMACHANDRULA, Sitaram , KUCHIBHOTLA, Anjaneyulu
发明人： RAMACHANDRULA, Sitaram , KUCHIBHOTLA, Anjaneyulu
IPC分类号： H04M3/493
CPC分类号： H04M3/4936 , H04M3/4938 , H04M2203/251 , H04M2203/254
摘要： The present invention provides a method and system for navigating and editing telephonic IVR information. Data fields are audibly presented from an IVR system to a telephone caller. Each data field is associated with a unique identifier. An edit mode is also audibly presented from the IVR system to the caller. During the edit mode, the caller can provide speech and/or touch-tone input corresponding to the unique identifiers respectively associated with the data fields to selectively, individually and non-sequentially: navigate data fields; audibly playback data fields; provide input to data fields; audibly playback input to data fields; modify input to data fields; confirm input to data fields.
摘要翻译：本发明提供一种用于导航和编辑电话IVR信息的方法和系统。数据字段从IVR系统可听见地呈现给电话呼叫者。每个数据字段与唯一标识符相关联。编辑模式也从IVR系统向主叫方提供。在编辑模式期间，呼叫者可以提供对应于分别与数据字段相关联的唯一标识符的语音和/或触摸音输入，以选择性地，单独地和非顺序地：导航数据字段; 可听回放数据字段; 为数据字段提供输入; 可视地播放输入到数据字段; 修改输入到数据字段; 确认输入到数据字段。

5. 发明申请

WO2012168942A1 IMAGE TRIGGERED TRANSACTIONS 审中-公开
标题翻译：图像触发交易
公开(公告)号：WO2012168942A1
公开(公告)日：2012-12-13
申请号：PCT/IN2011/000383
申请日：2011-06-08
申请人： HEWLETT-PACKARD DEVELOPMENT COMPANY , RAMACHANDRULA, Sitaram , MANDALAPU, Dinesh , KOMPALLI, Suryaprakash , KUCHIBHOTLA, Anjaneyulu Seetha Rama , MATAD, Nagabhusha Ayyanahal , GODAVARI, Srinivasu , MAJUNATH, Geetha
发明人： RAMACHANDRULA, Sitaram , MANDALAPU, Dinesh , KOMPALLI, Suryaprakash , KUCHIBHOTLA, Anjaneyulu Seetha Rama , MATAD, Nagabhusha Ayyanahal , GODAVARI, Srinivasu , MAJUNATH, Geetha
IPC分类号： G06K9/78 , H04N21/00
CPC分类号： G06F17/30247 , G06K9/00449
摘要： In one example, a method for image triggered transactions includes capturing an image of a document with an imaging device and searching a database for a template matching the captured image. If a matching template is found for the captured image, then retrieving pre-authored tasks associated with the captured image, extracting fields from the image related to the tasks, processing the extracted fields to obtain content to complete the tasks; and executing the selected task. If a matching template is not found for the captured image, then an authoring application is launched to add the captured image as a new template and author tasks associated with the captured image.
摘要翻译：在一个示例中，用于图像触发事务的方法包括利用成像设备捕获文档的图像，并且搜索与数据库匹配的捕获图像的数据库。如果为捕获的图像找到匹配的模板，则检索与捕获的图像相关联的预先编写的任务，从与任务相关的图像中提取字段，处理提取的字段以获得内容以完成任务; 并执行所选择的任务。如果没有为捕获的图像找到匹配的模板，则启动创作应用程序以将捕获的图像添加为新模板，并与捕获的图像相关联的作者任务。

6. 发明申请

WO2011001439A3 SKEW DETECTION 审中-公开
标题翻译： SKEW检测
公开(公告)号：WO2011001439A3
公开(公告)日：2011-08-11
申请号：PCT/IN2009000373
申请日：2009-07-02
申请人： HEWLETT PACKARD DEVELOPMENT CO , RAMACHANDRULA SITARAM , BANERJEE SERENE , PARIKH PULKIT , SHAFFI NOUSHATH , KUCHIBHOTLA ANJANEYULU SEETHA RAMA , DEY PRASENJIT
发明人： RAMACHANDRULA SITARAM , BANERJEE SERENE , PARIKH PULKIT , SHAFFI NOUSHATH , KUCHIBHOTLA ANJANEYULU SEETHA RAMA , DEY PRASENJIT
IPC分类号： G06T7/40 , G06K9/32 , G06K9/46
CPC分类号： H04N1/3878 , G06K9/3283 , G06K9/4633 , G06K9/4647 , G06K2209/01 , G06T7/12 , G06T7/194 , G06T7/70 , G06T2207/20061 , G06T2207/30176
摘要： Presented is a skew detection apparatus for estimating a skew angle of a document shown in an image. After a pre-processing step (100), skew is estimated by a first method which uses the alignment of pixels on the boundaries of the document and/or the content o the document (199). This may be an Adapted Quasi-Hough Transform (AQHT). If this first algorighm is not confident in this output, another skew esrimation approach is tried (299), wherein the second method relies in pixels in the interior rather than the boundary o the document content This second method may be Enhanced Piecewise Covering by Parallelograms (EPCP). If the second skew estimate has no sufficiency high confidence value, consistency between the first and second estimates is assessed (300).
摘要翻译：提出了一种用于估计图像中显示的文档的歪斜角度的偏斜检测装置。在预处理步骤（100）之后，通过使用文档的边界上的像素和/或文档（199）的内容的对齐的第一种方法来估计偏斜。这可能是适应的准霍夫变换（AQHT）。如果这个第一个算法在这个输出中没有信心，则尝试另一种偏斜约束方法（299），其中第二种方法依赖于内部的像素而不是文档内容的边界。第二种方法可以是通过平行四边形增强的分段覆盖 EPCP）。如果第二个歪斜估计没有足够的高置信度值，则评估第一和第二估计之间的一致性（300）。

你已经成功收藏专利！

检索式保存成功!

IPRDB

热门服务

关于我们

友情链接

联系方式