Probability estimate for K-nearest neighbor
    61.
    发明授权
    Probability estimate for K-nearest neighbor 有权
    K最近邻的概率估计

    公开(公告)号:US07451123B2

    公开(公告)日:2008-11-11

    申请号:US11296919

    申请日:2005-12-08

    IPC分类号: G06E1/00 G06F15/00

    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。

    Audio duplicate detector
    63.
    发明授权
    Audio duplicate detector 有权
    音频重复检测器

    公开(公告)号:US07421305B2

    公开(公告)日:2008-09-02

    申请号:US10785561

    申请日:2004-02-24

    IPC分类号: G06F17/00

    摘要: The present invention relates to a system and methodology to facilitate automatic management and pruning of audio files residing in a database. Audio fingerprinting is a powerful tool for identifying streaming or file-based audio, using a database of fingerprints. Duplicate detection identifies duplicate audio clips in a set, even if the clips differ in compression quality or duration. The present invention can be provided as a self-contained application that it does not require an external database of fingerprints. Also, a user interface provides various options for managing and pruning the audio files.

    摘要翻译: 本发明涉及一种便于自动管理和修剪驻留在数据库中的音频文件的系统和方法。 音频指纹是使用指纹数据库识别流媒体或基于文件的音频的强大工具。 重复的检测识别集合中的重复音频剪辑,即使剪辑在压缩质量或持续时间上有所不同。 本发明可以作为独立应用来提供,其不需要外部指纹数据库。 此外,用户界面提供了管理和修剪音频文件的各种选项。

    CONTEXT-BASED SEARCH, RETRIEVAL, AND AWARENESS
    66.
    发明申请
    CONTEXT-BASED SEARCH, RETRIEVAL, AND AWARENESS 审中-公开
    基于语境的搜索,检索和意识

    公开(公告)号:US20080005067A1

    公开(公告)日:2008-01-03

    申请号:US11426981

    申请日:2006-06-28

    IPC分类号: G06F17/30

    CPC分类号: G06F16/24575

    摘要: A system that incorporates a user context into a computer-based search is provided. To establish the context, the innovation can identify information about a user state or context via a variety of sources and sensors. The state/context information can be used to filter, arrange and/or rank search results so as to facilitate converging on meaningful searches and results. Machine learning systems (implicitly and/or explicitly trained) can be employed to infer a current and/or future context related to user. An identified or inferred user context can be employed to modify an automated or user-defined search input/query. Contextual cues can be considered directly in the construction and use of context of context-sensitive retrieval algorithms that are optimized for identifying and/or ranking of informational items of potential interest or value in different contexts. As well, the context can be employed to intelligently render results of a query (e.g., user/application defined, context-modified query).

    摘要翻译: 提供了将用户上下文合并到基于计算机的搜索中的系统。 为了建立上下文,创新可以通过各种来源和传感器识别关于用户状态或上下文的信息。 状态/上下文信息可用于过滤,排列和/或排列搜索结果,以便于有意义地搜索和结果的收敛。 可以使用机器学习系统(隐式和/或显式训练)来推断与用户相关的当前和/或将来的上下文。 可以使用识别或推断的用户上下文来修改自动或用户定义的搜索输入/查询。 上下文提示可以直接考虑在对不同上下文中潜在兴趣或价值的信息项目的识别和/或排名进行优化的上下文敏感检索算法的上下文的构建和使用中。 同样,可以使用上下文来智能地呈现查询的结果(例如,定义的用户/应用程序,上下文修改的查询)。

    Auto playlist generator
    68.
    发明授权
    Auto playlist generator 有权
    自动播放列表生成器

    公开(公告)号:US07313571B1

    公开(公告)日:2007-12-25

    申请号:US11263347

    申请日:2005-10-31

    IPC分类号: G06F17/30 G03B19/18

    摘要: A system and method for generating a list is provided. The system includes a seed item input subsystem, an item identifying subsystem, a descriptive metadata similarity determining subsystem and a list generating subsystem that builds a list based, at least in part, on similarity processing performed on seed item descriptive metadata and user item descriptive metadata and user selected thresholds applied to such similarity processing. The method includes inexact matching between identifying metadata associated with new user items and identifying metadata stored in a reference metadata database. The method further includes subjecting candidate user items to similarity processing, where the degree to which the candidate user items are similar to the seed item is determined, and placing user items in a list of items based on user selected preferences for (dis)similarity between items in the list and the seed item.

    摘要翻译: 提供了一种用于生成列表的系统和方法。 系统包括种子项目输入子系统,项目识别子系统,描述性元数据相似性确定子系统和列表生成子系统,其至少部分地基于对种子项目描述性元数据和用户项目描述元数据执行的相似性处理来构建列表 并且将用户选择的阈值应用于这种相似性处理。 该方法包括识别与新用户项目相关联的元数据和识别存储在参考元数据数据库中的元数据之间的不准确匹配。 该方法还包括对候选用户项目进行相似性处理,其中确定候选用户项目类似于种子项目的程度,并且基于用户选择的偏好来将用户项目放置在项目列表中以用于(dis)相似性 列表中的项目和种子项目。