公开(公告)号：US07016884B2

公开(公告)日：2006-03-21

申请号：US10183213

申请日：2002-06-27

申请人： John C. Platt ,  Christopher J. C. Burges

发明人： John C. Platt ,  Christopher J. C. Burges

IPC分类号： G06N3/02

CPC分类号： G06K9/6276

摘要： Systems and methods are disclosed that facilitate producing probabilistic outputs also referred to as posterior probabilities. The probabilistic outputs include an estimate of classification strength. The present invention intercepts non-probabilistic classifier output and applies a set of kernel models based on a softmax function to derive the desired probabilistic outputs. Such probabilistic outputs can be employed with handwriting recognition where the probability of a handwriting sample classification is combined with language models to make better classification decisions.

摘要翻译： 公开了有助于产生也称为后验概率的概率输出的系统和方法。 概率输出包括分类强度的估计。 本发明拦截非概率分类器输出并且基于softmax函数应用一组核心模型以导出所需的概率输出。 这样的概率输出可以与手写识别一起使用，其中手写样本分类的概率与语言模型组合以进行更好的分类决定。

公开(公告)号：US06992625B1

公开(公告)日：2006-01-31

申请号：US10423093

申请日：2003-04-25

申请人： John C. Krumm ,  John C. Platt

发明人： John C. Krumm ,  John C. Platt

IPC分类号： G01S3/02 ,  G01C17/00

CPC分类号： H04W64/00 ,  G01C21/206 ,  G01S5/0252 ,  H04W16/20 ,  H04W16/225 ,  H04W88/18

摘要： An architecture for minimizing calibration effort in an IEEE 802.11 device location measurement system. The calibration technique is based upon a regression function that produces adequately accurate location information as a function of signal strength regardless of gaps in the calibration data or minimally available data. The algorithm takes a set of signal strengths from known room locations in a building and generates a function giving (x,y) as a function of signal strength, which function may then be used for the estimation of new locations. Radial basis functions, which are simple to express and compute, are used for regression. The fact that the algorithm maps signal strength to continuous location makes it possible to skip rooms during calibration, yet still evaluate the location in those rooms.

摘要翻译： 用于最小化IEEE 802.11设备位置测量系统中的校准工作的架构。 校准技术基于回归函数，其产生与校准数据或最小可用数据中的间隙无关的信号强度的函数的足够精确的位置信息。 该算法从建筑物中的已知房间位置获取一组信号强度，并产生作为信号强度的函数给出（x，y），该函数然后可用于估计新位置。 用于表达和计算的简单的径向基函数用于回归。 该算法将信号强度映射到连续位置的事实使得可以在校准期间跳过房间，但仍然评估这些房间中的位置。

公开(公告)号：US06987221B2

公开(公告)日：2006-01-17

申请号：US10158674

申请日：2002-05-30

申请人： John C. Platt

发明人： John C. Platt

IPC分类号： G06F17/30

CPC分类号： G06F17/30758 ,  G06F17/30038 ,  G06F17/30053 ,  G06F17/30743 ,  G06F17/30749 ,  G06F17/30772 ,  G10H1/0058 ,  G10H2240/061 ,  G10H2240/245 ,  G10H2240/295 ,  Y10S707/99943 ,  Y10S707/99945

摘要： The present invention relates to systems and/or methods that generate playlist(s) for a library or collection of media items via selecting a plurality of seed items, at least one of which is an undesirable seed item. Some of the seed items are desirable indicating that a user prefers additional media items similar to the desirable seed items and others are undesirable indicating that the user prefers additional media items dissimilar to the undesirable seed items. Additionally, the seed items can be weighted to establish a relative importance of the seed items. The invention compares media items in the collection with the seed items and determines which media items are added into the playlist by computation of similarity metrics or values. The playlist can be regenerated by adding desirable seed items to the playlist and removing media items from the playlist (e.g., undesirable seed items).

公开(公告)号：US06973210B1

公开(公告)日：2005-12-06

申请号：US09481163

申请日：2000-01-12

申请人： John C. Platt ,  Donald P. Mitchell ,  J. Turner Whitted ,  James F. Blinn

发明人： John C. Platt ,  Donald P. Mitchell ,  J. Turner Whitted ,  James F. Blinn

IPC分类号： G06K9/00 ,  G06K9/40

CPC分类号： G09G3/20 ,  G09G2320/02 ,  G09G2340/0457

摘要： Image data processing and image rendering methods and systems whereby images are displayed on display devices having pixels with separately controllable pixel sub-components. Image data, such as data encoded in a three-channel signal, is passed through a low-pass filter to remove frequencies higher than a selected cutoff frequency, which obtain samples from the color components of the signal that map spatially different image regions to individual pixel sub-components. It has been found that color aliasing effects can be significantly reduces at a cutoff frequency somewhat higher than the Nyquist frequency, while enhancing the spatial resolution of the image. The image data is then pass through sampling filters, A generalized set of filters includes nine filters, one for each combination of one color and one pixel sub-component. The filtering coefficients of the filters can be selected to optimize of approximate an optimization of an error metric, which represents the color and luminance errors perceived on the display device. In this manner, a desired balance between color accuracy and luminance accuracy can be obtained. The samples mapped to individual pixel sub-components are used to generate luminous intensity values for the displayed image.

摘要翻译： 图像数据处理和图像渲染方法和系统，其中图像被显示在具有可分开控制的像素子组件的像素的显示装置上。 图像数据，例如以三通道信号编码的数据，通过低通滤波器以去除高于所选截止频率的频率，该频率从将空间不同图像区域映射到个体的信号的颜色分量获得样本 像素子组件。 已经发现，在提高图像的空间分辨率的截止频率下，颜色混叠效应可以显着地降低到比奈奎斯特频率稍高的截止频率。 图像数据然后通过采样滤波器。 一组广泛的滤波器包括九个滤波器，一个用于一个颜色和一个像素子组件的每个组合。 可以选择滤波器的滤波系数以优化误差度量的近似优化，其表示在显示装置上感知的颜色和亮度误差。 以这种方式，可以获得颜色精度和亮度精度之间的期望的平衡。 使用映射到各个像素子分量的样本来生成显示图像的亮度值。

公开(公告)号：US06933947B2

公开(公告)日：2005-08-23

申请号：US10308510

申请日：2002-12-03

申请人： John C. Platt ,  Mikhail M. Lyapunov

发明人： John C. Platt ,  Mikhail M. Lyapunov

IPC分类号： G06T1/00 ,  G06T5/00 ,  G06T11/20 ,  G09G5/00 ,  G09G5/36 ,  G09G5/377 ,  H04N5/202 ,  H04N5/66 ,  G09G5/02

CPC分类号： G06T11/203 ,  G06T15/503

摘要： The blending coefficients (alpha values) of font glyphs undergo alpha correction to compensate for a lack of gamma correction in text rendering processes. The alpha correction includes selecting a set of correction coefficients that correspond to the predetermined gamma value of the display device and computing corrected alpha values from the known alpha values, the foreground colors, and set of correction coefficients. The corrected alpha values can then be used to blend the foreground and background colors of the corresponding display pixels without requiring gamma correction. Accordingly, the alpha correction can be performed by a GPU, which is not configured to perform gamma correction, thereby increasing the speed at which text rendering can occur.

摘要翻译： 字体字形的混合系数（α值）经历alpha校正，以弥补文本呈现过程中缺少伽马校正。 α校正包括选择与显示装置的预定伽马值对应的一组校正系数，并根据已知的α值，前景色和校正系数组计算校正的α值。 然后可以使用校正的α值来混合相应显示像素的前景色和背景色，而不需要伽马校正。 因此，可以通过未被配置为执行伽马校正的GPU执行α校正，从而增加可以发生文本呈现的速度。

公开(公告)号：US06895112B2

公开(公告)日：2005-05-17

申请号：US09783505

申请日：2001-02-13

申请人： Tong-Xian Chen ,  Xiangrong Chen ,  John C. Platt ,  Jie Yan ,  Hong-Jiang Zhang

发明人： Tong-Xian Chen ,  Xiangrong Chen ,  John C. Platt ,  Jie Yan ,  Hong-Jiang Zhang

IPC分类号： G06K9/00 ,  G06T5/00 ,  G06T7/00 ,  H04N1/62

CPC分类号： G06K9/0061 ,  G06T7/0002 ,  G06T7/70 ,  G06T7/90 ,  G06T2207/30216 ,  H04N1/624

摘要： Red-eye detection based on red region detection with eye confirmation initially identifies pixels that correspond to the color of red-eye within an image. A determination is then made as to whether these identified pixels and surrounding areas are part of an eye or not part of an eye. Those identified pixels that are determined to be part of an eye are the detected red-eye regions.

公开(公告)号：US06694301B1

公开(公告)日：2004-02-17

申请号：US09540255

申请日：2000-03-31

申请人： David E. Heckerman ,  D. Maxwell Chickering ,  John C. Platt ,  Christopher A. Meek ,  Bo Thiesson

发明人： David E. Heckerman ,  D. Maxwell Chickering ,  John C. Platt ,  Christopher A. Meek ,  Bo Thiesson

IPC分类号： G06N502

CPC分类号： G06K9/6254 ,  G06K9/6226 ,  G06K9/6296 ,  G06N7/005

摘要： Clustering for purposes of data visualization and making predictions is disclosed. Embodiments of the invention are operable on a number of variables that have a predetermined representation. The variables include input-only variables, output-only variables, and both input-and-output variables. Embodiments of the invention generate a model that has a bottleneck architecture. The model includes a top layer of nodes of at least the input-only variables, one or more middle layer of hidden nodes, and a bottom layer of nodes of the output-only and the input-and-output variables. At least one cluster is determined from this model. The model can be a probabilistic neural network and/or a Bayesian network.

摘要翻译： 公开了用于数据可视化和进行预测的聚类。 本发明的实施例可以对具有预定表示的多个变量进行操作。 变量包括仅输入变量，仅输出变量，以及输入和输出变量。 本发明的实施例生成具有瓶颈架构的模型。 该模型包括至少仅输入变量，一个或多个中间层隐藏节点的顶层，以及仅输出和输入和输出变量的节点的底层。 从该模型确定至少一个群集。 该模型可以是概率神经网络和/或贝叶斯网络。

公开(公告)号：US06674436B1

公开(公告)日：2004-01-06

申请号：US09364649

申请日：1999-07-30

申请人： Bodin Dresevic ,  William Hill ,  Gregory Hitchcock ,  Leroy B. Keely ,  John C. Platt ,  J. Turner Whitted

发明人： Bodin Dresevic ,  William Hill ,  Gregory Hitchcock ,  Leroy B. Keely ,  John C. Platt ,  J. Turner Whitted

IPC分类号： G09G526

CPC分类号： G09G5/005 ,  G06F1/1607 ,  G09G5/00 ,  G09G5/006 ,  G09G5/02 ,  G09G2320/02 ,  G09G2320/0276 ,  G09G2320/028 ,  G09G2320/043 ,  G09G2320/06 ,  G09G2320/0606 ,  G09G2320/0626 ,  G09G2320/0666 ,  G09G2320/0673 ,  G09G2320/0693 ,  G09G2320/08 ,  G09G2340/0407 ,  G09G2340/0457 ,  G09G2360/144

摘要： The present invention relates to methods and apparatus for increasing the perceived quality of displayed images. This is achieved in a variety of ways including the use of a plurality of device specific display characteristics when preparing images for display. It is also achieved through the monitoring of display device and/or ambient light conditions, e.g., on a periodic basis, and using the obtained information when controlling display output. Another approach to improving the perceived quality of displayed images involves the use of information relating to a specific user's ability to perceive image characteristics such as color. By customizing display output to an individual user's own physical perception capabilities and/or viewing characteristics it is possible to enhance the image quality perceived by the individual viewer as compared to embodiments which do not take into consideration individual user characteristics.

摘要翻译： 本发明涉及增加显示图像感知质量的方法和装置。 这可以通过各种方式实现，包括在准备用于显示的图像时使用多个设备特定的显示特性。 还可以通过例如周期性地监视显示装置和/或环境光条件，并且在控制显示输出时使用获得的信息来实现。 提高显示图像的感知质量的另一种方法涉及使用与特定用户感知诸如颜色的图像特征的能力有关的信息。 通过将显示输出定制为单个用户自己的物理感知能力和/或观看特征，与不考虑个别用户特征的实施例相比，可以增强单个观看者感知到的图像质量。

公开(公告)号：US06393145B2

公开(公告)日：2002-05-21

申请号：US09364365

申请日：1999-07-30

申请人： Claude Betrisey ,  Bodin Dresevic ,  Donald P. Mitchell ,  John C. Platt

发明人： Claude Betrisey ,  Bodin Dresevic ,  Donald P. Mitchell ,  John C. Platt

IPC分类号： G06K900

CPC分类号： G09G5/005 ,  G09G5/006 ,  G09G5/024 ,  G09G5/28 ,  G09G2320/0242 ,  G09G2320/0276 ,  G09G2340/0457 ,  G09G2340/10

摘要： Techniques for improving the resolution of images (either analog images, analytic images, or images having a higher resolution than that of a display device) to be rendered on patterned displays. In one aspect of the present invention, an overscaling or oversampling process may accept analytic character information, such as contours for example, and a scale factor or grid and overscale or oversample the analytic character information to produce an overscaled or oversampled image. The overscaled or oversampled image generated has a higher resolution than the display upon which the character is to be rendered. Displaced samples of the overscaled or oversampled image are then combined (or filtered). An analytic image, such as a line drawing for example, may be applied to the oversampling/overscaling process as was the case with the character analytic image. However, since the analytic image may have different units than that of the character analytic image, the scale factor applied may be different. Since an ultra resolution image is already “digitized”, that is, not merely mathematically expressed contours or lines between points, it may be applied directly to a process for combining displaced samples of the ultra-resolution image to generate another ultra-resolution image (or an image with sub-pixel information). The functionality of the overscaling/oversampling process and the processes for combining displaced samples may be combined into a single step analytic to digital sub-pixel resolution conversion process.

摘要翻译： 用于提高在图案化显示器上呈现的图像（模拟图像，分析图像或具有比显示装置的分辨率更高的分辨率的图像）的分辨率的技术。 在本发明的一个方面，超标或过采样过程可以接受分析字符信息，例如轮廓，以及比例因子或网格，并且超分析或过采样分析字符信息以产生过度取样或过采样图像。 所生成的过大或过采样图像的分辨率要高于要显示字符的显示器。 然后组合过滤或过采样图像的位移样本（或过滤）。 分析图像，例如线图，可以应用于过采样/超量程过程，如字符分析图像的情况。 然而，由于分析图像可能具有与字符分析图像不同的单位，所以应用的比例因子可能不同。 由于超分辨率图像已经被“数字化”，即不仅仅是在数学上表达轮廓或点之间的线，所以它可以直接应用于用于组合超分辨率图像的位移样本以产生另一超分辨率图像的处理 或具有子像素信息的图像）。 过量程/过采样过程的功能和组合移位样本的过程可以组合成单步分析到数字子像素分辨率转换过程。

公开(公告)号：US6028959A

公开(公告)日：2000-02-22

申请号：US287590

申请日：1999-04-06

申请人： Chung-Ning Wang ,  John C. Platt ,  Nada P. Matic

发明人： Chung-Ning Wang ,  John C. Platt ,  Nada P. Matic

IPC分类号： G06K9/22 ,  G06K9/18

CPC分类号： G06K9/00436 ,  G06F3/018 ,  G06K9/00416 ,  G06K2209/011

摘要： A method for incremental recognition of ideographic handwriting comprises in order the steps of: (1) entering in a natural stroke order at least one stroke of an ideographic character from a computer entry tablet; (2) providing the at least one stroke to an incremental character recognizer, which produces a hypothesis list of at least one candidate character; (3) displaying a hypothesis list of candidate characters containing the at least one stroke; (4) selecting a correct character from among the candidate characters on the hypothesis list if it a correct character appears thereon; (5) entering in natural stroke order at least one additional stroke of the ideographic character from the computer entry tablet if no candidate character is a correct character; (6) providing the additional stroke(s) to the incremental character recognizer, which produces an updated hypothesis list; (7) displaying the updated hypothesis list of candidate characters containing every stroke; (8) selecting a correct character from among the candidate characters on the updated hypothesis list if it a correct character appears thereon; and (9) repeating steps (5) through (8) until a correct character is selected from the updated hypothesis list.

摘要翻译： 一种用于增量识别表意笔迹的方法包括以下步骤：（1）从计算机入口平板电脑输入至少一笔表意字符的自然笔画顺序; （2）向增量字符识别器提供至少一个笔划，其产生至少一个候选人物的假设列表; （3）显示包含所述至少一个笔画的候选字符的假设列表; （4）如果其中出现正确的字符，则从假设列表中的候选字符中选择正确的字符; （5）如果没有候选人物是正确的角色，则从计算机输入板输入自然笔划至少一个表意字符的附加笔画; （6）向所述增量字符识别器提供附加冲程，其产生更新的假设列表; （7）显示包含每个笔画的候选字符的更新假设列表; （8）如果在其上出现正确的字符，则从更新假设列表上的候选字符中选择正确的字符; 和（9）重复步骤（5）至（8），直到从更新的假设列表中选择正确的字符。