公开(公告)号：US20170019308A1

公开(公告)日：2017-01-19

申请号：US15042116

申请日：2016-02-11

Applicant: Netflix, Inc.

Inventor： Roy RAPOPORT ,  Christopher SANDEN ,  Cody RIOUX ,  Gregory BURRELL

IPC: H04L12/24 ,  H04L29/08

CPC classification number: H04L41/12 ,  G06F11/3409 ,  G06F11/3452 ,  H04L67/10

Abstract: One embodiment of the invention disclosed herein provides techniques for detecting and remediating an outlier server in a distributed computer system. A control server retrieves a group of time-series data sets associated with a first time period, where each time-series data set represents a performance metric for a different server in a group of servers. The control server generates a cluster that includes two or more of the time-series data sets, where the performance metric for each server that is associated with one of the time-series data sets in the cluster is within a threshold distance from the performance metric for the servers that are associated with the other time-series data sets in the cluster. The control server determines that a particular time-series data set corresponds to a server included in the group of servers and is not included in the cluster, and marks the server as an outlier server.

Abstract translation: 本文公开的本发明的一个实施例提供了用于检测和修复分布式计算机系统中的离群服务器的技术。 控制服务器检索与第一时间段相关联的一组时间序列数据集，其中每个时间序列数据集表示服务器组中的不同服务器的性能度量。 控制服务器生成包含两个或更多个时间序列数据集的集群，其中与集群中的一个时间序列数据集相关联的每个服务器的性能指标在与性能度量 对于与集群中其他时间序列数据集相关联的服务器。 控制服务器确定特定时间序列数据集对应于服务器组中包括的服务器，并且不包括在群集中，并将服务器标记为离群服务器。

公开(公告)号：US20160379057A1

公开(公告)日：2016-12-29

申请号：US14879045

申请日：2015-10-08

Applicant: NETFLIX, INC

Inventor： Ioannis KATSAVOUNIDIS

IPC: G06K9/00 ,  G06K9/46 ,  H04N9/64 ,  G06T7/40

CPC classification number: G06K9/00765 ,  G06K9/4652 ,  G06K9/52 ,  G11B27/28 ,  H04N9/64 ,  H04N21/23418 ,  H04N21/234363 ,  H04N21/2662 ,  H04N21/81

Abstract: In one embodiment of the present invention, a native resolution analyzer generates a log-magnitude spectrum that elucidates sampling operations that have been performed on a scene. In operation, the native resolution analyzer performs a transform operation of a color component associated with a frame included in the scene to generate a frame spectrum. The native resolution analyzer then normalizes the magnitudes associated with the frame spectrum and logarithmically scales the normalized magnitudes to create a log-magnitude frame spectrum. This two dimensional log-magnitude frame spectrum serves as a frequency signature for the frame. More specifically, patterns in the log-magnitude spectrum reflect re-sampling operations, such as a down-sampling and subsequent up-sampling, that may have been performed on the frame. By analyzing the log-magnitude spectrum, discrepancies between the display resolution of the scene and the lowest resolution with which the scene has been processed may be detected in an automated fashion.

Abstract translation: 在本发明的一个实施例中，原始分辨率分析器产生阐明已经在场景上执行的采样操作的对数幅度谱。 在操作中，原始分辨率分析器执行与包括在场景中的帧相关联的颜色分量的变换操作以产生帧频谱。 然后，原始分辨率分析器对与帧频谱相关联的幅度进行归一化，并且对数地缩放归一化幅度以产生对数幅度帧频谱。 该二维对数幅度帧频谱用作帧的频率特征。 更具体地，对数幅度频谱中的图案反映可能已经在帧上执行的重采样操作，例如下采样和随后的上采样。 通过分析对数幅度谱，可以以自动方式检测场景的显示分辨率与已经处理场景的最低分辨率之间的差异。

公开(公告)号：US20160335754A1

公开(公告)日：2016-11-17

申请号：US14709230

申请日：2015-05-11

Applicant: NETFLIX, Inc

Inventor： Anne Aaron ,  Dae Kim ,  Yu-Chieh Lin ,  David Ronca ,  Andy Schuler ,  Kuyen Tsao ,  Chi-Hao Wu

IPC: G06T7/00 ,  G06T7/20

CPC classification number: G06T7/0002 ,  G06T7/20 ,  G06T9/002 ,  G06T2207/10016 ,  G06T2207/20084 ,  G06T2207/30168 ,  H04N19/154 ,  H04N21/466 ,  H04N21/4666

Abstract: In one embodiment of the present invention, a quality trainer and quality calculator collaborate to establish a consistent perceptual quality metric via machine learning. In a training phase, the quality trainer leverages machine intelligence techniques to create a perceptual quality model that combines objective metrics to optimally track a subjective metric assigned during viewings of training videos. Subsequently, the quality calculator applies the perceptual quality model to values for the objective metrics for a target video, thereby generating a perceptual quality score for the target video. In this fashion, the perceptual quality model judiciously fuses the objective metrics for the target video based on the visual feedback processed during the training phase. Since the contribution of each objective metric to the perceptual quality score is determined based on empirical data, the perceptual quality score is a more accurate assessment of observed video quality than conventional objective metrics.

Abstract translation: 在本发明的一个实施例中，质量培训者和质量计算器通过机器学习协作建立一致的知觉质量度量。 在培训阶段，质量培训师利用机器智能技术创建感知质量模型，结合客观指标，最佳地跟踪在观看培训视频期间分配的主观度量。 随后，质量计算器将感知质量模型应用于目标视频的客观指标的值，从而生成目标视频的感知质量得分。 以这种方式，感知质量模型基于在训练阶段处理的视觉反馈，明智地融合了目标视频的客观指标。 由于每个客观度量对知觉质量得分的贡献是基于经验数据确定的，感知质量得分是比传统客观指标更准确地评估观察到的视频质量。

公开(公告)号：US20160292158A1

公开(公告)日：2016-10-06

申请号：US15079767

申请日：2016-03-24

Applicant: NETFLIX, INC.

Inventor： Michael Todd ,  Ehtsham Elahi ,  Aish Fenton ,  Carlos Gomez Uribe

IPC: G06F17/30 ,  H04L29/08 ,  H04L29/06

Abstract: A computer-implemented method and a system for implementation the method are provided. The method includes receiving user activity data for a first set of users of a streaming media system, in which the first set of users having access to a first catalog. Additional user activity data for a second set of users is received. The second set of users has access to a second catalog, which includes media items also present in the first catalog. The second catalog further includes an exclusive media item that is not included in the first catalog. A probability that a user of the first set of users would play the exclusive media item if presented by the streaming media system with the exclusive media item is determined. Information is provided to a user of the second set of users as a recommendation to play the exclusive media item based on the probability.

Abstract translation: 提供了一种计算机实现的方法和用于实现该方法的系统。 该方法包括接收流媒体系统的第一组用户的用户活动数据，其中第一组用户可访问第一目录。 接收第二组用户的附加用户活动数据。 第二组用户可以访问第二个目录，其中包括也存在于第一个目录中的媒体项目。 第二个目录还包括不包含在第一个目录中的独家媒体项目。 如果由具有专用媒体项目的流媒体系统呈现，则确定第一组用户的用户将播放专用媒体项的概率。 基于该概率，向第二组用户的用户提供信息作为播放专属媒体项目的推荐。

公开(公告)号：US09454530B2

公开(公告)日：2016-09-27

申请号：US13644548

申请日：2012-10-04

Applicant: NETFLIX Inc.

Inventor： Mohammad Sabah ,  Carlos Gomez-Uribe ,  Vijay Bharadwaj ,  Sasi Parthasarathy ,  Siddharth Angrish

IPC: G06F7/00 ,  G06F17/30

CPC classification number: G06F17/30029

Abstract: Techniques are described for determining relationships between user activities and determining search results and content recommendations based on the relationships. A plays-related-to-searches application may determine a relationship score between plays of a media title and searches of a query by determining a distance between a projection of the search onto the space of the users and a projection of plays of the media title onto the space of the users. A plays-after-searches application may determine a score for plays of the streaming media title given the search by multiplying a number of times plays of the media title occur after the query is entered by the number of times any play occurs, and dividing by a product of the number of times plays of the media title occur after any query is entered and the number of times plays of any media title occur after the query is entered.

Abstract translation: 描述了用于确定用户活动之间的关系并基于关系确定搜索结果和内容推荐的技术。 戏剧相关搜索应用可以通过确定搜索的投影到用户的空间和媒体标题的播放的投影之间的距离来确定媒体标题的播放与查询的搜索之间的关系得分 到用户的空间。 播放后搜索应用程序可以通过乘以在查询输入之后出现的媒体标题的播放次数乘以发生任何播放的次数，并且除以 媒体标题的播放次数的产物在输入任何查询后发生，并且在输入查询后发生任何媒体标题的播放次数。

公开(公告)号：US09430212B2

公开(公告)日：2016-08-30

申请号：US14177182

申请日：2014-02-10

Applicant: Netflix, Inc.

Inventor： Michael Tripoli ,  Brian Moyles

IPC: G06F9/455 ,  G06F9/45 ,  G06F9/445 ,  H04L29/08

CPC classification number: G06F8/63 ,  G06F9/45533 ,  G06F2009/45562 ,  G06F2009/4557 ,  H04L67/1095

Abstract: In an embodiment, a method comprises using a first server computer, executing a process that is configured to generate virtual computer instances; receiving a request to generate a first custom virtual computer instance, wherein the request identifies one or more software packages to install on the first custom virtual computer instance; copying a first pre-configured volume image onto a volume of a data storage device, wherein the volume includes one or more base software packages in a file system; mounting the volume to the first server computer, in response to receiving the request; installing each of the one or more software packages on the volume; unmounting the volume; creating a snapshot of the volume; registering the snapshot as a registered custom volume image.

Abstract translation: 在一个实施例中，一种方法包括使用第一服务器计算机，执行被配置为生成虚拟计算机实例的进程; 接收生成第一自定义虚拟计算机实例的请求，其中所述请求标识要在所述第一自定义虚拟计算机实例上安装的一个或多个软件包; 将第一预先配置的卷映像复制到数据存储设备的卷上，其中所述卷包括文件系统中的一个或多个基本软件包; 响应于接收到所述请求而将所述卷安装到所述第一服务器计算机; 在卷上安装一个或多个软件包的每个; 卸载音量; 创建卷的快照; 将快照注册为注册的自定义卷映像。

公开(公告)号：US20160247252A1

公开(公告)日：2016-08-25

申请号：US15049986

申请日：2016-02-22

Applicant: Netflix, Inc.

Inventor： MATTHEW SEELEY ,  KIM TROTT ,  MATTHEW MCCARTHY ,  DYLAN OUDYK ,  SAM MAGNUSON ,  JAN HANSSEN

IPC: G06T1/60 ,  G06F12/08 ,  G06F17/30

CPC classification number: G06F12/0875 ,  G06F12/0207 ,  G06F12/123 ,  G06F2212/455 ,  H04N19/423 ,  H04N19/46

Abstract: In an embodiment, a data processing method provides an improvement in efficient use of computer memory and comprises using a computer, creating in computer memory a glyph memory area that is configured to store a plurality of cached glyphs; using the computer, receiving a request from an application to use a particular glyph; in response to the request, determining whether the particular glyph is in the glyph memory area; in response to determining that the particular glyph is not in the glyph memory area: attempting to store a bitmap of the particular glyph to a next location in the glyph memory area; in response to determining that the next location is not available a first time, reclaiming space in the glyph memory area in an amount sufficient to store the bitmap; attempting a second time to store the bitmap in the next location in the glyph memory area; in response to determining that the next location is not available a second time, clearing the glyph memory area of all previously stored glyphs and storing the bitmap in the glyph memory area.

Abstract translation: 在一个实施例中，数据处理方法提供了计算机存储器的有效使用的改进，并且包括使用计算机，在计算机存储器中创建被配置为存储多个缓存字形的字形存储器区域; 使用计算机，从应用程序接收使用特定字形的请求; 响应于所述请求，确定所述特定字形是否在所述字形存储器区域中; 响应于确定特定字形不在字形存储器区域中：尝试将特定字形的位图存储到字形存储器区域中的下一个位置; 响应于第一次确定下一位置不可用，以足以存储位图的量来回收字形存储器区域中的空间; 尝试第二次将位图存储在字形存储器区域中的下一个位置; 响应于第二次确定下一位置不可用，清除所有先前存储的字形的字形存储器区域并将位图存储在字形存储器区域中。

公开(公告)号：US20160072870A1

公开(公告)日：2016-03-10

申请号：US14942084

申请日：2015-11-16

Applicant: Netflix, Inc.

Inventor： Mark Watson ,  Bin Wang ,  Wei Wei

IPC: H04L29/06

Abstract: Techniques for transmitting a digital content stream through a network between a plurality of content servers and a content player. Network performance information for the plurality of content servers is retrieved. Embodiments then determine a mapping of the plurality of content servers to a plurality of portions of a content buffer on the content player, based on the retrieved network performance information. A request is transmitted to each of the plurality of content servers to transmit digital content associated with the corresponding portion of the content buffer.

Abstract translation: 用于通过多个内容服务器和内容播放器之间的网络发送数字内容流的技术。 检索多个内容服务器的网络性能信息。 然后，实施例基于所检索的网络性能信息来确定多个内容服务器与内容播放器上的内容缓冲区的多个部分的映射。 向多个内容服务器中的每一个发送请求以发送与内容缓冲器的相应部分相关联的数字内容。

公开(公告)号：US20160057461A1

公开(公告)日：2016-02-25

申请号：US14832990

申请日：2015-08-21

Applicant: NETFLIX, INC.

Inventor： Christopher Stewart JAFFE ,  Rico NASOL ,  Michael BELL ,  John MCLORD ,  Joubert NEL

IPC: H04N21/234 ,  H04N21/2662 ,  H04N21/81 ,  H04N21/258

CPC classification number: H04N21/44222 ,  G06F3/04812 ,  G06F3/0482 ,  H04N21/23418 ,  H04N21/23424 ,  H04N21/2353 ,  H04N21/2387 ,  H04N21/25891 ,  H04N21/2662 ,  H04N21/4312 ,  H04N21/44 ,  H04N21/44008 ,  H04N21/44218 ,  H04N21/4532 ,  H04N21/47217 ,  H04N21/812 ,  H04N21/8456

Abstract: One embodiment of the present invention sets forth a technique for displaying one or more video streams on a consumer device. The technique includes setting a first display characteristic associated with a first video stream to a first value based on a first priority associated with the first video stream. The technique further includes transmitting the first video stream to a consumer device for display. The technique further includes selecting a second video stream, setting a second display characteristic associated with the second video stream to a second value based on a second priority associated with the second video stream, and transmitting the second video stream to the consumer device for display.

Abstract translation: 本发明的一个实施例提出了一种用于在消费者设备上显示一个或多个视频流的技术。 该技术包括基于与第一视频流相关联的第一优先级，将与第一视频流相关联的第一显示特性设置为第一值。 该技术还包括将第一视频流发送到消费者设备进行显示。 该技术还包括：基于与第二视频流相关联的第二优先级，选择第二视频流，将与第二视频流相关联的第二显示特性设置为第二值，以及将第二视频流发送到消费者设备进行显示。

公开(公告)号：US20160042015A1

公开(公告)日：2016-02-11

申请号：US14454632

申请日：2014-08-07

Applicant: NETFLIX, INC.

Inventor： AIRAN LANDAU ,  SUDHIR TONSE

IPC: G06F17/30 ,  G06F9/54

Abstract: Techniques for activity information schema discovery, schema change detection, and notification. In one embodiment, for example, a computer-implemented method for activity information schema discovery and schema change detection and notification comprises: analyzing a first set of related activity messages obtained during a first sample period; determining first schema counters for uniquely named properties identified in the first set of messages based on the analyzing of the first set of activity messages; after the first sample period, inferring a first schema from the first schema counters; analyzing a second set of related activity messages obtained during a second sample period; determining second schema counters for uniquely named properties identified in the second set of messages based on the analyzing of the second set of activity messages; after the second sample period, inferring a second schema from the second schema counters; comparing the first schema and the second schema for any differences.

Abstract translation: 用于活动信息模式发现，模式更改检测和通知的技术。 在一个实施例中，例如，用于活动信息模式发现和模式改变检测和通知的计算机实现的方法包括：分析在第一采样周期期间获得的第一组相关活动消息; 基于对所述第一组活动消息的分析，确定在所述第一组消息中标识的唯一命名属性的第一模式计数器; 在第一个采样周期之后，从第一个模式计数器推断第一个模式; 分析在第二采样周期期间获得的第二组相关活动消息; 基于对所述第二组活动消息的分析，确定在所述第二组消息中标识的唯一命名属性的第二模式计数器; 在第二个采样周期之后，从第二个模式计数器推断第二个模式; 比较第一个模式和第二个模式的任何差异。