International Business Machines Corporation

Globalfoundries Inc.

Wendy Chong

Nirmit V Desai

Heather D Achilles

Shahrokh Daijavad

Alberto Valdes Garcia

Jean-Olivier Plouchart

Petros Zerfos

Steven Edward Millman

Carmelo I Uria

James R Kozloski

Dinesh C Verma

Jenny S Li

Alejandro Gabriel Schrott

Levente Klein

Janani Janakiraman

Anuradha Rao

Joefon Jann

Catherine Helen Crawford

John Jeremy Rice

Bijan Davari

Pablo Meyer Rojas

Tai-chi Su

Ronald C Geiger, Jr

Joe Latone

Kelvin Kakugawa

Prach Jerry Chuaypradit

Peter F Mycue

Singpui Zee

Adam E Grant

Randy M Fussell

Christopher P Huff

Kelvin Kakugwa

Personalization with on-device privacy preservation is provided by receiving, at a user device, a generalized recommendation from a remote provider backend, which include a plurality of candidate items responsive to a request from an application running on the user device; generating a plurality of candidate vectors corresponding to the plurality of candidate items representing a corresponding candidate item in a binary format across a plurality of metrics associated with the plurality of candidate items; generating a user vector based on personal data stored on the user device, the user vector representing a user in a binary format across one or more metrics associated with the personal data; determining interaction propensities based on the user vector and the plurality of candidate vectors according to a global model run on the user device; and displaying, on the user device, the plurality of candidate items according to the interaction propensities.

On-device privacy-preservation and personalization

A notification message is received by a communication device The notification message includes a message identifier and a digital signature generated by an original sender of the notification message. That an integrity of message content in the notification message and an original sender of the notification message is verified is determined using the digital signature. Responsive to determining that the notification message is verified, the notification message is forwarded by the communication device using a peer-to-peer communication channel over a mesh network.

Delivering messages to offline devices using peer-to-peer communication

A trained base model is distributed to a set of nodes. From a first node in the set of nodes, a first set of meta-metrics resulting from a transfer learning operation on the trained base model at the first node is collected. The transfer learning at the first node uses first local data available at the first node. The first node is clustered in a cluster with a second node from the set of nodes, in response to a meta-metric in the first set of meta-metrics being within a tolerance value of a corresponding meta-metric in a second set of meta-metrics collected from the second node. A normalized set of model parameters is constructed after an iteration of transfer learning or local learning at the first and second nodes. The normalized set of model parameters is distributed to the first node and the second node in the cluster.

Transfer learning without local data export in multi-node machine learning

Methods and systems for tag-based identification include receiving a set of parameters at a user device from a remote server. A counterfeit-proof identification tag is read using a sensor in the user device using the set of parameters. Features of the counterfeit-proof identification tag are extracted in accordance with a feature extraction function, using a processor, to generate a tag bit sequence. A challenge function is applied to the extracted features to generate a result. The result is transmitted to the remote server to authenticate the counterfeit-proof identification tag. The counterfeit-proof identification tag is authenticated with a tag database at the remote server.

Consumer and business anti-counterfeiting services using identification tags

Methods and systems for tag-based identification include reading a counterfeit-proof identification tag using a sensor in a user device. Features of the identification tag are extracted in accordance with a feature extraction function, using a processor, to generate a tag bit sequence. A challenge function is applied to the extracted features to generate a result. The result is transmitted to a remote server to authenticate the identification tag.

Various embodiments provide a customized cosmetics recommendation for a specific user. In one embodiment a method comprises capturing an image that includes the face of the specific user, producing a set of hyperspectral images from the image, analyzing the hyperspectral images to determine a set of spectral components of the face, and providing a recommendation for one or more cosmetics customized for the specific user based on the set of spectral components and cosmetician expert judgement. The image may be captured using a hyperspectral imaging camera. The set of spectral components is compared to a plurality of previous sets of spectral components to find a match and one or more cosmetics mapped to the match are provided as the recommendation. Additionally, a set of conditional options may be received and one or more cosmetics mapped to the set of conditional options and the set of spectral components are provided as the recommendation.

Method system and medium for personalized expert cosmetics recommendation using hyperspectral imaging

To mitigate collisions in a physical space during gaming, a set of physical objects and a user situated in the 3D space are mapped to determine a spacing between an object in the set of physical objects and the user, where the user moves in the 3D space to cause a motion in a virtual environment of a game. A prediction is computed that the user will make a motion in the 3D space during the gaming. For the motion in the 3D space, a motion trajectory of the user is computed using a measurement parameter corresponding to the user stored in a user profile. A detection is made that the motion trajectory of the user violates a spacing threshold between the user and the object, and the user is alerted about a risk of collision between the user and the object in the 3D space.

Mitigating collisions in a physical space during gaming

Growing community of inventors

Carmel, NY, United States of America

Wendy Chong