公开(公告)号：US20240095593A1

公开(公告)日：2024-03-21

申请号：US18368209

申请日：2023-09-14

申请人： Irdeto B.V.

发明人： Robert DURAND ,  Philip EISEN ,  Thomas HICKIE

IPC分类号： G06N20/00

CPC分类号： G06N20/00

摘要： A machine learning model protection method comprising: generating, based on a set of parameters that define a machine learning model, an item of software which, when executed by one or more processors, provides an implementation for the machine learning model; and applying one or more software protection techniques to the item of software.

公开(公告)号：US20230214484A1

公开(公告)日：2023-07-06

申请号：US18089273

申请日：2022-12-27

申请人： Irdeto B.V.

发明人： Thomas HICKIE ,  Robert DURAND

IPC分类号： G06F21/55 ,  G06F8/41

CPC分类号： G06F21/554 ,  G06F8/427 ,  G06F2221/033

摘要： Systems, methods, and storage media for creating secured computer code are disclosed. Exemplary implementations may: access computer code; convert the computer code into a numeric description of characteristics of the code; partition the computer code into blocks of code; determine a corresponding ranking of at least some of the blocks of code with an anomaly measure by applying an anomaly detection algorithm to the blocks of code; select anomalous blocks of the blocks of code by applying a threshold to the rankings; and apply code security techniques to at least one of the anomalous blocks of code to thereby create secured computer code.

公开(公告)号：US20230196195A1

公开(公告)日：2023-06-22

申请号：US17982896

申请日：2022-11-08

申请人： IRDETO B.V.

发明人： Thomas HICKIE ,  Shufei HE

IPC分类号： G06N20/00

CPC分类号： G06N20/00

摘要： A method for identifying whether a classification system is configured to use a specific machine-learning classification model, the method comprising: using the classification system to generate, for each test sample in a predetermined test set that comprises a plurality of test samples, a corresponding classification result; and identifying either (i) that the classification system is using the specific machine-learning classification model if, for each test sample in the test set, the corresponding classification result matches a classification result produced for that test sample using the specific machine-learning classification model or (ii) that the classification system is not using the specific machine-learning classification model if there is a test sample in the test set for which the corresponding classification result does not match the classification result produced for that test sample using the specific machine-learning classification model; wherein the test set is associated with the specific machine-learning classification model and, for each test sample in the test set, there is a corresponding small modification for that test sample that causes a change in the classification result produced for that test sample using the specific machine-learning classification model.

公开(公告)号：US20240303328A1

公开(公告)日：2024-09-12

申请号：US18442402

申请日：2024-02-15

申请人： Irdeto B.V.

发明人： Thomas HICKIE

IPC分类号： G06F21/55 ,  G06N5/04 ,  G06N20/00

CPC分类号： G06F21/554 ,  G06N5/04 ,  G06N20/00 ,  G06F2221/033

摘要： Disclosed implementations include a method of detecting attacks on Machine Learning (ML) models by applying the concept of anomaly detection based on the internal state of the model being protected. Instead of looking at the input or output data directly, disclosed implementation look at the internal state of the hidden layers of a neural network of the model after processing of data. By examining how different layers within a neural network model are behaving an inference can be made as to whether the data that produced the observed state is anomalous (and thus possibly part of an attack on the model).