公开(公告)号：US11954140B2

公开(公告)日：2024-04-09

申请号：US17666388

申请日：2022-02-07

申请人： Verint Systems Ltd.

发明人： Roni Romano

IPC分类号： G06F16/36 ,  G06F40/247 ,  G06F40/30 ,  G06N5/022

CPC分类号： G06F16/367 ,  G06F40/247 ,  G06F40/30 ,  G06N5/022

摘要： By formulizing a specific company's internal knowledge and terminology, the ontology programming accounts for linguistic meaning to surface relevant and important content for analysis. The ontology is built on the premise that meaningful terms are detected in the corpus and then classified according to specific semantic concepts, or entities. Once the main terms are defined, direct relations or linkages can be formed between these terms and their associated entities. Then, the relations are grouped into themes, which are groups or abstracts that contain synonymous relations. The disclosed ontology programming adapts to the language used in a specific domain, including linguistic patterns and properties, such as word order, relationships between terms, and syntactical variations. The ontology programming automatically trains itself to understand the domain or environment of the communication data by processing and analyzing a defined corpus of communication data.

公开(公告)号：US20240021206A1

公开(公告)日：2024-01-18

申请号：US18475599

申请日：2023-09-27

申请人： VERINT SYSTEMS INC.

发明人： Omer Ziv ,  Ran Achituv ,  Ido Shapira ,  Jeremie Dreyfuss

IPC分类号： G10L17/00 ,  G10L17/02

CPC分类号： G10L17/00 ,  G10L17/02

摘要： Systems and method of diarization of audio files use an acoustic voiceprint model. A plurality of audio files are analyzed to arrive at an acoustic voiceprint model associated to an identified speaker. Metadata associate with an audio file is used to select an acoustic voiceprint model. The selected acoustic voiceprint model is applied in a diarization to identify audio data of the identified speaker.

公开(公告)号：US08805009B1

公开(公告)日：2014-08-12

申请号：US13351493

申请日：2012-01-17

申请人： Shahar Daliyot ,  Gadi Zimerman ,  Shmuel Kiro ,  Oren Golan

发明人： Shahar Daliyot ,  Gadi Zimerman ,  Shmuel Kiro ,  Oren Golan

IPC分类号： G06K9/00

CPC分类号： G06K9/6255

摘要： Embodiments disclosed herein provide systems and methods for matching trajectories across disjointed video views. In a particular embodiment, a method provides identifying a first trajectory associated with an object in video captured of a first region. The method further provides identifying a plurality of alternative views of the object from the video of the first region. The method further provides using the plurality of alternative views of the object, selecting a second trajectory that corresponds to the first trajectory from a plurality of trajectories associated with a plurality of objects in video captured of a second region.

摘要翻译： 本文公开的实施例提供了用于在不相交的视频视图上匹配轨迹的系统和方法。 在特定实施例中，一种方法提供识别与第一区域捕获的视频中的对象相关联的第一轨迹。 该方法还提供从第一区域的视频识别对象的多个替代视图。 该方法还提供使用对象的多个替代视图，从与第二区域的视频捕获的多个对象相关联的多个轨迹中选择对应于第一轨迹的第二轨迹。

公开(公告)号：US08781289B2

公开(公告)日：2014-07-15

申请号：US13589611

申请日：2012-08-20

申请人： Hing Yip Chung ,  Fuk Sang Mak ,  Golan Levy ,  Wai Chung Lam ,  Chong Va Cheong ,  Shiu Hang Tsang

发明人： Hing Yip Chung ,  Fuk Sang Mak ,  Golan Levy ,  Wai Chung Lam ,  Chong Va Cheong ,  Shiu Hang Tsang

IPC分类号： H04N5/92

CPC分类号： H04N5/7605 ,  H04N5/765 ,  H04N5/77 ,  H04N7/181 ,  H04N7/185 ,  H04N21/2747

摘要： Embodiments disclosed herein provide systems and methods for performing video recorder failover. In a particular embodiment, a system includes a first Network Video Recorder (NVR) configured to receive a video stream and record the video stream to a first long-term storage. The system further includes a second NVR configured to receive the video stream and temporarily stores an amount of the video stream in temporary storage. In response to a detection of a failure of the first NVR, the second NVR is configured to record the video stream to the second long-term storage and transfer at least a portion of the video stream stored in the temporary storage corresponding to the amount of time between detection of the failure and when the second NVR began recording the video stream to the second long-term storage.

摘要翻译： 本文公开的实施例提供了用于执行录像机故障切换的系统和方法。 在特定实施例中，系统包括被配置为接收视频流并将视频流记录到第一长期存储器的第一网络视频记录器（NVR）。 该系统还包括被配置为接收视频流并将视频流的量临时存储在临时存储中的第二NVR。 响应于对第一NVR的故障的检测，第二NVR被配置为将视频流记录到第二长期存储并且传送存储在临时存储器中的视频流的至少一部分，对应于 检测到故障与第二个NVR开始将视频流录制到第二个长期存储器之间的时间。

公开(公告)号：US20130215267A1

公开(公告)日：2013-08-22

申请号：US13856178

申请日：2013-04-03

申请人： Verint Systems Inc.

发明人： Oren Golan ,  Boaz Dudovich ,  Shahar Daliyot ,  Itzik Horovitz ,  Shmuel Kiro

IPC分类号： H04N7/18

CPC分类号： H04N7/181 ,  G06K9/00771 ,  G06K9/6253 ,  G06Q30/02 ,  G06T7/70 ,  H04N7/18

摘要： A system and method for the capture and storage of data relating to the movements of objects, in a specified area and enables this data to be displayed in a graphically meaningful and useful manner. Video data is collected and video metadata is generated relating to objects (persons) appearing in the video data and their movements over time. The movements of the objects are then analyzed to detect the movements within a region of interest. This detection of movement allows a user, such as a manager of a store, to make informed decisions as to the infrastructure and operation of the store. One detection method relates to the number of people that are present in a region of interest for a specified time period. A second detection method relates to the number of people that remain or dwell in a particular area for a particular time period. A third detection method determines the flow of people and the direction they take within a region of interest. A fourth detection method relates to the number of people that enter a certain area by crossing a virtual line, a tripwire.

公开(公告)号：US08249413B1

公开(公告)日：2012-08-21

申请号：US12045543

申请日：2008-03-10

申请人： Hing Yip Chung ,  Fuk Sang Mak ,  Golan Levy ,  Wai Chung Lam ,  Chong Va Cheong ,  Shiu Hang Tsang

发明人： Hing Yip Chung ,  Fuk Sang Mak ,  Golan Levy ,  Wai Chung Lam ,  Chong Va Cheong ,  Shiu Hang Tsang

IPC分类号： H04N5/00

CPC分类号： H04N5/7605 ,  H04N5/765 ,  H04N5/77 ,  H04N7/181 ,  H04N7/185 ,  H04N21/2747

摘要： A video recording system includes a first Network Video Recorder (NVR) that receives a video delivered via a video packet stream and records the video to a first long-term storage. A second NVR receives the same video and temporarily stores it in temporary storage. When a failure of the first NVR is detected at a first point, the second NVR starts recording the video to long-term storage. The temporarily stored video is used by the second NVR to fill in the gap in recorded video in long-term storage that occurred between the failure of the first NVR and the start of the second NVR recording the video to long-term storage.

摘要翻译： 视频记录系统包括第一网络视频记录器（NVR），其接收通过视频分组流传送的视频并将视频记录到第一长期存储器。 第二个NVR接收相同的视频，并将其临时存储在临时存储中。 当在第一点检测到第一NVR的故障时，第二NVR开始将视频记录到长期存储。 临时存储的视频被第二NVR用于填补在第一NVR的故障和第二NVR开始将视频记录到长期存储之间发生的长期存储中的记录视频中的间隙。

公开(公告)号：US07907142B2

公开(公告)日：2011-03-15

申请号：US12014155

申请日：2008-01-15

申请人： Marc Racicot

发明人： Marc Racicot

IPC分类号： G06F15/16 ,  G09G5/00

CPC分类号： H04N5/44591 ,  H04N7/0806 ,  H04N7/12 ,  H04N7/181 ,  H04N19/61 ,  H04N19/85 ,  H04N21/4263 ,  H04N21/4314 ,  H04N21/440263

摘要： A system and method for creating video tile arrays is disclosed. A first plurality of video streams produced by a first plurality of Analog to Digital Converters (ADCs) are received in a first Digital Signal Processor (DSP). The first DSP processes the first plurality of video streams to produce a partial video tile array stream. The partial video tile array stream is transferred to a second DSP. The second DSP receives a second plurality of video streams produced by a second plurality of video ADCs. The second DSP processes the partial video tile array stream and the second plurality of video streams to generate a tiled video stream. The tiled video stream comprises an array of a plurality of video tiles that each correspond to a one of the first and second plurality of video streams.

摘要翻译： 公开了一种用于创建视频瓦片阵列的系统和方法。 在第一数字信号处理器（DSP）中接收由第一多个模数转换器（ADC）产生的第一多个视频流。 第一DSP处理第一多个视频流以产生部分视频瓦片阵列流。 部分视频瓦片阵列流被传送到第二DSP。 第二DSP接收由第二多个视频ADC产生的第二多个视频流。 第二DSP处理部分视频图块阵列流和第二多个视频流以生成平铺视频流。 平铺视频流包括多个视频瓦片的阵列，每个视频瓦片对应于第一和第二多个视频流中的一个。

公开(公告)号：US20100013931A1

公开(公告)日：2010-01-21

申请号：US12504492

申请日：2009-07-16

申请人： Oren Golan ,  Boaz Dudovich ,  Shahar Daliyot ,  Itzik Horovitz ,  Shmuel Kiro

发明人： Oren Golan ,  Boaz Dudovich ,  Shahar Daliyot ,  Itzik Horovitz ,  Shmuel Kiro

IPC分类号： H04N7/18

CPC分类号： H04N7/181 ,  G06K9/00771 ,  G06K9/6253 ,  G06Q30/02 ,  G06T7/70 ,  H04N7/18

摘要： A system and method for the capture and storage of data relating to the movements of objects, in a specified area and enables this data to be displayed in a graphically meaningful and useful manner. Video data is collected and video metadata is generated relating to objects (persons) appearing in the video data and their movements over time. The movements of the objects are then analyzed to detect the movements within a region of interest. This detection of movement allows a user, such as a manager of a store, to make informed decisions as to the infrastructure and operation of the store. One detection method relates to the number of people that are present in a region of interest for a specified time period. A second detection method relates to the number of people that remain or dwell in a particular area for a particular time period. A third detection method determines the flow of people and the direction they take within a region of interest. A fourth detection method relates to the number of people that enter a certain area by crossing a virtual line, a tripwire.

摘要翻译： 一种用于捕获和存储与指定区域中的对象移动有关的数据的系统和方法，并使得该数据能以图形有意义和有用的方式显示。 收集视频数据并生成与视频数据中出现的对象（人物）及其随时间的移动相关的视频元数据。 然后分析对象的移动以检测感兴趣区域内的移动。 这种移动检测允许诸如商店的管理者之类的用户对商店的基础设施和操作做出明智的决定。 一种检测方法涉及在指定时间段内存在于感兴趣区域中的人的数量。 第二种检测方法涉及在特定时间段内在特定地区保留或居住的人数。 第三种检测方法决定了人们的流动以及他们在感兴趣的区域内的方向。 第四种检测方法涉及通过穿过虚拟线路（tripwire）进入某一区域的人数。

公开(公告)号：US09503674B2

公开(公告)日：2016-11-22

申请号：US14942189

申请日：2015-11-16

申请人： Verint Systems Inc.

发明人： Hing Yip Chung ,  Fuk Sang Mak ,  Golan Levy ,  Wai Chung Lam ,  Chong Va Cheong ,  Shiu Hang Tsang

IPC分类号： H04L29/06 ,  G06F21/00 ,  H04N5/76 ,  H04N5/765 ,  H04N5/77 ,  H04N21/2747 ,  H04N7/18

CPC分类号： H04N5/7605 ,  H04N5/765 ,  H04N5/77 ,  H04N7/181 ,  H04N7/185 ,  H04N21/2747

摘要： Embodiments disclosed herein provide systems and methods for performing video recorder failover. In a particular embodiment, a system for handing a failover of a first Network Video Recorder (NVR) is provided. The system includes a second NVR that receives a video stream and temporarily stores an amount of the video stream to the temporary storage, wherein the amount of the video stream stored in the temporary storage at any given time corresponds to a duration of time sufficient to accommodate a failover of the first NVR to the second NVR. In response to a detection of a failure of the first NVR, the second NVR records the video stream to the second long-term storage and transfers at least a portion of the video stream stored in the temporary storage to the second long-term storage.

摘要翻译： 本文公开的实施例提供了用于执行录像机故障切换的系统和方法。 在特定实施例中，提供了用于处理第一网络录像机（NVR）的故障转移的系统。 该系统包括第二NVR，其接收视频流并临时将一定量的视频流存储到临时存储器中，其中在任何给定时间存储在临时存储器中的视频流的量对应于足以容纳的持续时间 第一个NVR到第二个NVR的故障转移。 响应于检测到第一NVR的故障，第二NVR将视频流记录到第二长期存储器，并将存储在临时存储器中的视频流的至少一部分传送到第二长期存储器。

公开(公告)号：US09268996B1

公开(公告)日：2016-02-23

申请号：US13355285

申请日：2012-01-20

申请人： Gadi Zimerman ,  Itshak Horovitz ,  Shmuel Akerman ,  Amir Sole

发明人： Gadi Zimerman ,  Itshak Horovitz ,  Shmuel Akerman ,  Amir Sole

IPC分类号： G06K9/00

CPC分类号： G06K9/00711 ,  G06K9/00342 ,  G06K9/00348 ,  G06K9/00369 ,  G06K9/46 ,  G06K9/4671 ,  G06K9/6262 ,  G06T7/251 ,  G06T2207/10016 ,  G06T2207/30196 ,  G06T2207/30241

摘要： Models are generated from objects identified in video. Each model is evaluated based on knowledge of the objects determined from video analysis, and preferred models are identified based on the evaluations. In some examples, each model could be evaluated by tracking a movement of each object in the video by using each model to track the object from which it was generated, evaluating an ability of each model to identify the objects in the video that are similar to the object from which it was generated, and determining an amount of false identifications made by each model of different objects in different video that does not include the object from which it was generated.

摘要翻译： 模型是从视频中识别的对象生成的。 基于从视频分析确定的对象的知识来评估每个模型，并且基于评估来确定优选模型。 在一些示例中，可以通过使用每个模型来跟踪视频中的每个对象的运动来跟踪其生成的对象来评估每个模型的评估，评估每个模型识别视频中的对象的能力，类似于 从其生成的对象，并且确定不同视频中的不包括生成对象的不同对象的每个模型所做的虚假标识的量。