Nxp B.v.

Intrinsic Id B.v.

Christiaan Kuipers

Joppe Willem Bos

Pim Theo Tuyls

Nikita Veshchikov

Christine Van Vredendaal

Bruce Murray

Simon Johann Friedberger

Brian Ermans

Helmut Alexander Goettl

Sven Heine

Vincent Verneuil

A method and data processing system for making a machine learning model more resistant to adversarial examples are provided. In the method, an input for a machine learning model is provided. A randomly generated mask is added to the input to produce a modified input. The modified input is provided to the machine learning model. The randomly generated mask negates the effect of a perturbation added to the input for causing the input to be an adversarial example. The method may be implemented using the data processing system.

Method and data processing system for making machine learning model more resistent to adversarial examples

A data processing system having a PUF and method for providing multiple enrollments, or instantiations, of the PUF are provided. A PUF segment includes a plurality of SRAM cells on an integrated circuit. A PUF response from the PUF segment is used to create a first activation code and a first PUF key. A second PUF key may be created from the PUF response. Initially, during a second enrollment, the PUF response is combined with the first activation code to reproduce a codeword. The first secret string is reconstructed by encoding the codeword. The codeword is combined with the first activation code to reproduce the PUF response. Inverse anti-aging is applied to the PUF response. Then a second secret string is generated using a random number generator (RNG). The second secret string is encoded to produce a new codeword. The new codeword is combined with the recovered PUF response to create a second activation code. The second activation coded is hashed with the second secret string to provide a second PUF key.

Method for performing multiple enrollments of a physically uncloneable function

A distribution system and method for distributing digital information is provided, which has high recoverability from a security breach. The distribution system comprises a server () and a computing device (). During an enrollment phase, the computing device obtains a first response from an integrated physically unclonable function () integrated in the computing device. The system comprises an enrollment module () for determining helper data from a decryption key and the first response to enable later reconstruction of the decryption key from the helper data and a second response obtained from the physically unclonable function. During a reconstruction phase, which occurs after the enrollment phase and typically after a security breach has occurred that revealed data and/or programming code of the computing device, the server may encrypt digital information using an encryption module () with a cryptographic encryption key corresponding to the decryption key.

Distribution system and method for distributing digital information

Growing community of inventors

Eindhoven, Netherlands

Christiaan Kuipers