Porsche Automobil Holding has patented a method for detecting adversarial perturbations in neural network input data. The system utilizes a conditional generative adversarial network, comprising a generator and discriminator, to identify and flag compromised data, enhancing the reliability of automated or assisted vehicle operations. GlobalData’s report on Porsche Automobil Holding gives a 360-degree view of the company including its patenting strategy. Buy the report here.
Smarter leaders trust GlobalData
According to GlobalData’s company profile on Porsche Automobil Holding, was a key innovation area identified from patents. Porsche Automobil Holding's grant share as of July 2024 was 52%. Grant share is based on the ratio of number of grants to total number of patents.
Adversarial perturbation detection for neural networks in vehicles
The granted patent US12073329B2 outlines a system designed to enhance the reliability of automated or assisted transportation vehicle operations by detecting adversarial perturbations in input data. Central to this system is a backend server that trains a conditional generative adversarial network (CGAN), which consists of a generator and a discriminator network. During the training phase, the generator creates adversarial perturbations based on data from transportation vehicle sensors, while the discriminator is trained to identify these perturbations. In the application phase, the trained discriminator evaluates incoming data before it is processed by a neural network, flagging any adversarially perturbed data and adjusting the confidence levels associated with the sensor data and the sensor itself.
Additionally, the patent describes a detection device that operates in conjunction with the backend server. This device utilizes the trained discriminator network to assess input data for adversarial perturbations during the application phase. The detection results inform the confidence levels of both the input data and the output generated by the neural network, which is crucial for functions such as environment monitoring and automated driving. The system also allows for further training phases to adapt to new adversarial perturbations and includes provisions for testing defense strategies against such perturbations. Overall, the patent presents a comprehensive approach to improving the robustness of neural networks used in transportation systems by systematically addressing the challenges posed by adversarial attacks.
To know more about GlobalData’s detailed insights on Porsche Automobil Holding, buy the report here.
Data Insights
From
The gold standard of business intelligence.
Blending expert knowledge with cutting-edge technology, GlobalData’s unrivalled proprietary data will enable you to decode what’s happening in your market. You can make better informed decisions and gain a future-proof advantage over your competitors.
GlobalData, the leading provider of industry intelligence, provided the underlying data, research, and analysis used to produce this article.
GlobalData Patent Analytics tracks bibliographic data, legal events data, point in time patent ownerships, and backward and forward citations from global patenting offices. Textual analysis and official patent classifications are used to group patents into key thematic areas and link them to specific companies across the world’s largest industries.
