公开(公告)号：WO2019182725A1

公开(公告)日：2019-09-26

申请号：PCT/US2019/019353

申请日：2019-02-25

Applicant: PINDROP SECURITY, INC.

Inventor： GUPTA, Payas ,  NELMS, Terry

IPC: H04W12/06 ,  G10L15/30 ,  H04M3/42 ,  G06F21/32 ,  G06F21/31

Abstract: Disclosed herein are embodiments of systems, methods, and products comprises an authentication server for authentication leveraging multiple audio channels. The server receives an authentication request regarding a user upon the user interacting with a first electronic device. The server requests the first device to transmit a first audio file of an audio sample to the server. The audio sample may be the user's audio command or a machine-generated audio signal. The server requests a second electronic device to transmit a second audio file that is the recording of the same audio sample to the server. The second electronic device is a trusted device in proximity of the first device and executes an authentication function to enable the recording and transmitting of the audio sample. The server determines a similarity score between the first audio file and the second audio file and authenticates the user based on the similarity score.

公开(公告)号：WO2018148298A1

公开(公告)日：2018-08-16

申请号：PCT/US2018/017249

申请日：2018-02-07

Applicant: PINDROP SECURITY, INC.

Inventor： KHOURY, Elie ,  GARLAND, Matthew

IPC: G06F21/32 ,  G10L17/04

Abstract: A score indicating a likelihood that a first subject is the same as a second subject may be calibrated to compensate for aging of the first subject between samples of age-sensitive biometric characteristics. Age of the first subject obtained at a first sample time and age of the second subject obtained at a second sample time may be averaged, and an age approximation may be generated based on at least the age average and an interval between the first and second samples. The age approximation, the interval between the first and second sample times, and an obtained gender of the subject are used to calibrate the likelihood score.

公开(公告)号：WO2018027138A1

公开(公告)日：2018-02-08

申请号：PCT/US2017/045506

申请日：2017-08-04

Applicant: PINDROP SECURITY, INC.

Inventor： STRONG, Scott ,  PATIL, Kailash ,  DEWEY, David ,  BANDYOPADHYAY, Raj ,  CALHOUN, Telvis ,  BALASUBRAMANIYAN, Vijay

IPC: H04M3/493 ,  G06F21/50 ,  G06F21/32 ,  G06F21/55 ,  G06N99/00 ,  H04M15/00 ,  H04M3/527 ,  H04M7/00 ,  H04W12/12

Abstract: Systems and methods for call detail record (CDR) analysis to determine a risk score for a call and identify fraudulent activity and for fraud detection in Interactive Voice Response (IVR) systems. An example method may store information extracted from received calls. Queries of the stored information may be performed to select data using keys, wherein each key relates to one of the received calls, and wherein the queries are parallelized. The selected data may be transformed into feature vectors, wherein each feature vector relates to one of the received calls and includes a velocity feature and at least one of a behavior feature or a reputation feature. A risk score for the call may be generated during the call based on the feature vectors.

Abstract translation: 用于呼叫详细记录（CDR）分析的系统和方法，用于确定呼叫的风险评分并识别欺诈活动以及交互式语音响应（IVR）系统中的欺诈检测。 示例方法可以存储从接收到的呼叫中提取的信息。 可以执行所存储的信息的查询以使用密钥来选择数据，其中每个密钥涉及所接收的呼叫中的一个，并且其中所述查询是并行的。 所选数据可被变换成特征向量，其中每个特征向量与所接收到的呼叫之一相关并且包括速度特征以及行为特征或信誉特征中的至少一个。 根据特征向量，呼叫期间可能会产生呼叫风险分数。

公开(公告)号：WO2018026912A1

公开(公告)日：2018-02-08

申请号：PCT/US2017/045090

申请日：2017-08-02

Applicant: PINDROP SECURITY, INC.

Inventor： DOUGLAS, Lance

IPC: H04M3/22 ,  H04M3/436 ,  H04L29/06 ,  H04M7/00

Abstract: Aspects of the invention determining a threat score of a call traversing a telecommunications network by leveraging the signaling used to originate, propagate and terminate the call. Outer-edge data utilized to originate the call may be analyzed against historical, or third party real-time data to determine the propensity of calls originating from those facilities to be categorized as a threat. Storing the outer edge data before the call is sent over the communications network permits such data to be preserved and not subjected to manipulations during traversal of the communications network. This allows identification of threat attempts based on the outer edge data from origination facilities, thereby allowing isolation of a compromised network facility that may or may not be known to be compromised by its respective network owner. Other aspects utilize inner edge data from an intermediate node of the communications network which may be analyzed against other inner edge data from other intermediate nodes and/or outer edge data.

Abstract translation: 本发明的各方面通过利用用于发起，传播和终止呼叫的信令来确定穿过电信网络的呼叫的威胁评分。 可以根据历史或第三方实时数据分析用于发起呼叫的外部边缘数据，以确定来自这些设施的呼叫的倾向被归类为威胁。 在通过通信网络发送呼叫之前存储外部边缘数据允许保留这些数据，并且在遍历通信网络期间不会受到操纵。 这允许基于来自始发设施的外部边缘数据识别威胁尝试，从而允许隔离可能被或可能不被其相应的网络所有者入侵的被破坏的网络设施。 其他方面利用来自通信网络的中间节点的内部边缘数据，其可以相对于来自其他中间节点和/或外部边缘数据的其他内部边缘数据进行分析。

公开(公告)号：WO2017066648A1

公开(公告)日：2017-04-20

申请号：PCT/US2016/057154

申请日：2016-10-14

Applicant: PINDROP SECURITY, INC.

Inventor： BANDYOPADHYAY, Raj ,  CALHOUN, Telvis ,  BALASUBRAMANIYAN, Vijay ,  STRONG, Scott ,  PATIL, Kailash ,  DEWEY, David

IPC: H04M3/493 ,  G06F21/50

CPC classification number: H04M3/527 ,  G06F21/32 ,  G06F21/552 ,  G06N99/005 ,  H04M3/493 ,  H04M7/0078 ,  H04M15/41 ,  H04M2203/551 ,  H04M2203/6027 ,  H04W12/12

Abstract: Systems and methods for call detail record (CDR) analysis to determine a risk score for a call and identify fraudulent activity and for fraud detection in Interactive Voice Response (IVR) systems. An example method may store information extracted from received calls. Queries of the stored information may be performed to select data using keys, wherein each key relates to one of the received calls, and wherein the queries are parallelized. The selected data may be transformed into feature vectors, wherein each feature vector relates to one of the received calls and includes a velocity feature and at least one of a behavior feature or a reputation feature. A risk score for the call may be generated during the call based on the feature vectors.

Abstract translation: 用于呼叫详细记录（CDR）分析的系统和方法，用于确定呼叫的风险评分并识别欺诈活动以及交互式语音响应（IVR）系统中的欺诈检测。 示例方法可以存储从接收到的呼叫中提取的信息。 可以执行所存储的信息的查询以使用密钥来选择数据，其中每个密钥涉及所接收的呼叫中的一个，并且其中所述查询是并行的。 所选数据可被变换成特征向量，其中每个特征向量与所接收到的呼叫之一相关并且包括速度特征以及行为特征或信誉特征中的至少一个。 根据特征向量，呼叫期间可能会产生呼叫风险分数。

公开(公告)号：ES2993990T3

公开(公告)日：2025-01-15

申请号：ES18712052

申请日：2018-03-02

Applicant: PINDROP SECURITY INC

Inventor： KHOURY ELIE ,  NAGARSHETH PARAV ,  PATIL KAILASH ,  GARLAND MATTHEW

IPC: G10L17/02 ,  G10L17/00 ,  G10L17/04 ,  G10L17/06 ,  G10L17/18 ,  G10L25/30 ,  G10L25/51

Abstract: Un sistema de verificación automática de hablantes (ASV) incorpora una primera red neuronal profunda para extraer características acústicas profundas, como características CQCC profundas, de una muestra de voz recibida. Las características acústicas profundas son procesadas por una segunda red neuronal profunda que clasifica las características acústicas profundas según una probabilidad determinada de incluir una condición de suplantación. A continuación, un clasificador binario clasifica la muestra de voz como genuina o suplantada. (Traducción automática con Google Translate, sin valor legal)

公开(公告)号：AU2021254787A1

公开(公告)日：2022-10-27

申请号：AU2021254787

申请日：2021-04-15

Applicant: PINDROP SECURITY INC

Inventor： KHOURY ELIE ,  SIVARAMAN GANESH ,  KUMAR AVROSH ,  ANTOLIC-SOBAN IVAN

IPC: G10L17/00 ,  G06F21/32 ,  G10L15/22 ,  G10L17/04 ,  G10L17/06

Abstract: Embodiments described herein provide for a voice biometrics system execute machine-learning architectures capable of passive, active, continuous, or static operations, or a combination thereof. Systems passively and/or continuously, in some cases in addition to actively and/or statically, enrolling speakers as the speakers speak into or around an edge device (e.g., car, television, radio, phone). The system identifies users on the fly without requiring a new speaker to mirror prompted utterances for reconfiguring operations. The system manages speaker profiles as speakers provide utterances to the system. Machine-learning architectures implement a passive and continuous voice biometrics system, possibly without knowledge of speaker identities. The system creates identities in an unsupervised manner, sometimes passively enrolling and recognizing known or unknown speakers. The system offers personalization and security across a wide range of applications, including media content for over-the-top services and IoT devices (e.g., personal assistants, vehicles), and call centers.

公开(公告)号：AU2021277642A1

公开(公告)日：2021-12-23

申请号：AU2021277642

申请日：2021-11-30

Applicant: PINDROP SECURITY INC

Inventor： KHOURY ELIE ,  NAGARSHETH PARAV ,  PATIL KAILASH ,  GARLAND MATTHEW

IPC: G10L25/51 ,  G10L17/00 ,  G10L25/30

Abstract: Abstract An automated speaker verification (ASV) system incorporates a first deep neural network to extract deep acoustic features, such as deep CQCC features, from a received voice sample. The deep acoustic features are processed by a second deep neural network that classifies the deep acoustic features according to a determined likelihood of including a spoofing condition. A binary classifier then classifies the voice sample as being genuine or spoofed.

公开(公告)号：AU2017305245B2

公开(公告)日：2021-12-16

申请号：AU2017305245

申请日：2017-08-01

Applicant: PINDROP SECURITY INC

Inventor： GAUBITCH NICK ,  STRONG SCOTT ,  CORNWELL JOHN ,  KINGRAVI HASSAN ,  DEWEY DAVID

IPC: H04M3/22 ,  H04M7/12 ,  H04Q1/45

Abstract: Systems, methods, and computer-readable media for call classification and for training a model for call classification, an example method comprising: receiving DTMF information from a plurality of calls; determining, for each of the calls, a feature vector including statistics based on DTMF information such as DTMF residual signal comprising channel noise and additive noise; training a model for classification; comparing a new call feature vector to the model; predicting a device type and geographic location based on the comparison of the new call feature vector to the model; classifying the call as spoofed or genuine; and authenticating a call or altering an IVR call flow.

公开(公告)号：CA3036533C

公开(公告)日：2020-04-21

申请号：CA3036533

申请日：2017-09-11

Applicant: PINDROP SECURITY INC

Inventor： KHOURY ELIE ,  GARLAND MATTHEW

IPC: G10L17/04 ,  G10L17/18

Abstract: The present invention is directed to a deep neural network (DNN) having a triplet network architecture, which is suitable to perform speaker recognition. In particular, the DNN includes three feed-forward neural networks, which are trained according to a batch process utilizing a cohort set of negative training samples. After each batch of training samples is processed, the DNN may be trained according to a loss function, e.g., utilizing a cosine measure of similarity between respective samples, along with positive and negative margins, to provide a robust representation of voiceprints.