Patent · US Active

Neural network trained using federated learning with local training data preserved at local edge circuits

US12072954B1 · kind B1 · utility

2Cited by

2References

9Claims

0Family size

Assignee

NVIDIA Corporation · US

Inventors

Wenqi Li · Kunming, CN
Fausto Milletari · Munich, DE
Daguang Xu · Princeton, US
Yan Cheng · Beijing, CN
Nicola Christin Rieke · Munich, DE
Charles Jonathan Hancox · Chidden, GB
Wentao Zhu · Beijing, CN
Rong-Shyang Ou · Palo Alto, US
Andrew Feng · Cupertino, US

Key dates

Filing date	Nov 6, 2019
Grant date	Aug 27, 2024
Priority date	—
Expiry date	Feb 27, 2041

Classification

Technology area (CPC G)Physics
CPC primaryG06V2201/03
WIPO fieldComputer technology
WIPO sectorElectrical engineering

Abstract

Apparatuses, systems, and techniques to perform federated training of neural networks while maintaining control over dissemination of local models of neural networks from which aspects of local training data might be extracted. In at least one embodiment, a neural network is trained on local training data and a local model is provided to be aggregated with other local models into a global model that is in turn used for further local model training, wherein a provided local model or training is adjusted to reduce an ability to extract aspects of local training data therefrom.

Source: USPTO / EPO open patent data. Objective bibliographic and citation counts.