Visa International Service Association

Sunpreet Singh Arora

Kim R Wagner

Lacey Best-Rowden

John F Sheets

Hao Yang

Gaven James Watson

Srinivasan Raghuraman

Yuhang Wu

Saman Sarraf

Saikrishna Badrinarayanan

Maliheh Shirvanian

Sunit Lohtia

Saloni Vijaykumar Mahajan

Dennis Franklin Olson

Sneha Sri Tadepalli

Binoy Parag Parikh

Jorge Andre Torres Perez Palacios

Amisha Sisodiya

Jiayun He

Andres Jimenez

Amy Dawson

Elianna Starr

Sowmya Vuddaraju

Mihai Christodorescu

Mastooreh Salajegheh

Chunxi Jiang

Luan Nguyen

Mengjun Leng

Yangyong Zhang

Yizhen Wang

Hao Ngo

Shashank Agrawal

Ke Wang

Brian Sullivan

Dinah Sloan

Jason Blackhurst

Yanhong Wu

Sikhar Patranabis

Andrew Beams

Harshal Bhadreshkumar Shah

Panagiotis Chatzigiannis

Sebastian Meiser

Ahmed Abusnaina

Jason Lightman

Cori Christine Tymoszek

Methods and systems for performing federated private anomaly detection are disclosed. An anomaly detector computer can collaborate with an aggregator computer and an account management computer in order to train machine learning models, which can be used to classify events as not fraudulent or fraudulent. The anomaly detector computer can obliviously use private information (e.g., account flags and account flag values) held by the account management computer to train and use the machine learning models, such that the anomaly detector does not become aware of the account flags or account flags values. Such a system can be used, for example, for the detection and prevention of financial crime. The anomaly detector computer can use the account flag information possessed by the account management computer to identify fraudulent events performed by customers of the organization operating the account management computer.

Secure anomaly detection using federated learning

In some embodiments, a malware detection system includes an attack channel removal unit, a feature extraction unit coupled to the attack channel removal unit, and a graphical encoding unit coupled to the feature extraction unit and a malware detection unit. In some embodiments, based upon graphically-encoded component-based features and monotonic features extracted from attack-channel-free software output by the attack channel removal unit, the malware detection unit detects malware in software input into the malware detection system. In some embodiments, the monotonic features extracted from the attack-channel free software and the graphically-encoded component-based features are combined to generate a combination monotonic-component based feature vector. In some embodiments, the combination monotonic-component based feature vector is used to detect malware using the malware detection system.

Malware detection and mitigation system and method therefor

A method includes receiving a first encrypted first identity attribute. A first doubly encrypted first identity attribute is formed by encrypting the first encrypted first identity attribute. A second doubly encrypted first identity attribute is formed by encrypting the first encrypted first identity attribute. They are transmitted to a user device, which removes a user layer of encryption on each to form a second encrypted first identity attribute and a third encrypted first identity attribute. Layers of encryption are added to the second encrypted first identity attribute to form a third doubly encrypted first identity attribute and the third encrypted first identity attribute to form a fourth doubly encrypted first identity attribute. The server computer receives them and transmits, to the second identity provider computer, the fourth doubly encrypted first identity attribute. The second identity provider computer obtains a first identity attribute and compares it to a second identity attribute.

Secure data exchange matching across identity providers

Provided is a computer-implemented method for authenticating an identification document. The method includes determining, with at least one processor, whether image data associated with the identification document has at least one predetermined indicia. In response to determining that the image data has the at least one predetermined indicia, the method includes determining whether the at least one predetermined indicia corresponds to at least one invalidation mark on the identification document, and, in response to determining that the at least one predetermined indicia corresponds to the at least one invalidation mark, determining, that the identification document is invalid. In response to determining that the identification document is invalid, the method includes preventing or causing the prevention of at least one action from being performed. A system and computer program product for authenticating identification documents are also disclosed.

System, method, and computer program product for authenticating identification documents

A method includes a server computer receiving, from a first data provider computer, encrypted data derived from first identity data and a cryptographic key or derivative thereof stored at the first data provider computer. The server computer transmits, to a second data provider computer, the encrypted data and/or the cryptographic key or derivative thereof. The server computer receives, from the second data provider computer, intermediate data derived from second identity data stored at the second data provider computer. The server computer determines if the first identity data and the second identity data are duplicates while the first identity data and the second identity data are encrypted. The server computer removes one of encrypted first identity data, derived from the first identity data, and encrypted second identity data, derived from the second identity data, from a memory in the server computer.

Privacy-preserving data deduplication

Provided is a computer-implemented method for authenticating a customer during payment transactions based on biometric identification parameters of the customer that includes receiving image data associated with an image template for identification of a customer, receiving image data associated with an image of a biometric identification parameter of the customer during a payment transaction between the customer and a merchant, establishing a short-range communication connection with a user device associated with the customer during the payment transaction between the customer and the merchant, authenticating an identity of the customer for the payment transaction via the short-range communication connection, determining an account identifier of an account of the customer based on authenticating the identity of the customer for the payment transaction, and processing the payment transaction using the account identifier of the account of the customer. A system and computer program product are also disclosed.

Authentication based on biometric identification parameter of an individual for payment transaction

A method is disclosed. The method includes receiving, by a processing network computer from a relying party computer associated with a relying party, a request for data associated with a user operating a user device. The processing network computer may retrieve first encrypted data of the user having a user-layer of encryption. The processing computer can then generate a second symmetric key to add a relying party-layer of encryption to the first encrypted data using a stream cipher. The doubly encrypted data may be transmitted to a user device that removes the user-layer of encryption on the first doubly encrypted data, and then adds a second relying party-layer of encryption to form second doubly encrypted data. The second doubly encrypted data may be transmitted to the relying party computer, which can remove both relying party-layers of encryption to gain access to the data associated with the user.

Privacy preserving identity data exchange based on hybrid encryption

A delegated biometric authentication system and related methods are disclosed. Using the system, a user can securely delegate biometric authentication to a public device from his communication device. This public device may be an Internet of things device that is not owned by the user, such as a computer, smart TV, tablet, etc. The public device may operate in a public place, such as a hotel or library. The communication device may be the users own smartphone or tablet, etc. A fuzzy vault process can be used to store the user's biometric template in the system Embodiments preserver the user's privacy without compromising authentication security and user convenience.

Delegated biometric authentication

Provided are systems for determining adversarial examples that include at least one processor to determine a first additional input from a plurality of additional inputs based on a proximity of the first additional input to an initial input, determine a second additional input from the plurality of additional inputs based on a proximity of the second additional input to the first additional input, generate a first vector embedding, a second vector embedding and a third vector embedding based on the second additional input, generate a first relational embedding, a second relational embedding, and a third relational embedding based on the third vector embedding and the first vector embedding, concatenate the first relational embedding, the second relational embedding, and the third relational embedding to provide a concatenated version, and determine whether the first input is an adversarial example based on the concatenated version. Methods and computer program products are also provided.

System, method, and computer program product for determining adversarial examples

Training an adversarial perturbation detector comprises accessing a training set comprising an enrolled biometric sample xand a public biometric sample x of an enrolled user, and submitted biometric samples x' of a second user, the submitted biometric samples x′ comprising perturbed adversarial samples x′+Δx′. A transformation function k(⋅) is provided having learnable a parameter θ and a classifier having a learnable parameter σ. The training set is used to learn the parameters θ and σ by inputting the training set to the transformation function k(⋅). The transformation function k(⋅) generates transformed enrolled samples k(x), a transformed public biometric sample k(x), and a transformed adversarial sample k(x′+Δx′). The classifier classifies the transformed adversarial sample k(x′+Δx′) as a success or as a fail based on the transformed enrolled samples k(x). Based on a result of the classification, the learnable parameters θ and σ are updated.

Using an enrolled biometric dataset to detect adversarial examples in biometrics-based authentication system

Methods and systems for privacy-preserving identity attribute verification are presented. During an interaction between a relying entity and a user, a relying entity computer can transmit a policy token to a user device. The policy token may indicate the information needed by the relying entity in order to perform the interaction. The user device can verify the policy token, then use the policy token in conjunction with an identity token to generate a zero-knowledge proof. The user device may transmit the zero-knowledge proof to an identity service provider computer. The identity service provider computer may verify the zero-knowledge proof, then generate a verification message. The identity service provider computer may sign the verification message and transmit the signed verification message to the relying entity computer. The relying entity computer may verify the verification message and complete the interaction with the user.

Privacy-preserving identity attribute verification using policy tokens

Described herein are a system and techniques for detecting whether biometric data provided in an access request is genuine or a replay. In some embodiments, the system uses an machine learning model trained using genuine and replay sample data which is optimized in order to produce a result set in which results for the genuine samples are pulled closer to a genuine center and results for the replay samples are pushed away from the genuine center. Subjecting input biometric data (e.g., an audio sample) to the trained model results in a classification of the input biometric data as genuine or replay, which can then be used to determine whether or not to verify the input biometric data.

Replay spoofing detection for automatic speaker verification system

A method is disclosed. The method comprises receiving, by an identity network computer, a query set including a plurality of test identity attributes. After receiving the query set, the identity network computer may retrieve derivatives of identity attributes associated with a user, and an encrypted trapdoor, then compute an obscured query set using the query set, and optionally the derivatives of identity attributes. The identity network computer may transmit the obscured query set (i) and the encrypted trapdoor to a user device associated with the user, which generates and transmits a first modified trapdoor and the obscured query set to a relying party computer, or (ii) and a second modified trapdoor to the relying party computer. The relying party computer may thereafter use the obscured query set, and the first modified trapdoor or the second modified trapdoor, to determine if the identity attributes is a member of the query set.

Integrating identity tokens and privacy-preserving identity attribute attestations into interactions

Growing community of inventors

San Mateo, CA, United States of America

Sunpreet Singh Arora