eIQ® Model Watermarking Technology

EIQ-WATERM-TECH

Collecting data and training ML models are substantial investments and data ownership is a differentiator. However, a copyist can typically copy or clone a model with little effort. Besides extracting a model from a device via a memory dump, attacks have been published showing that to clone a model, access to the input-output behavior of the ML algorithm already suffices. Furthermore, copyright and patent laws are not applicable directly to most ML model IP.

To help developers in adding a piece of copyright-protected information to an ML model as to strengthen the copyright claim, and to provide a means to prove unauthorized copying, NXP has developed a method for embedding a watermark in a model. This feature is available within the eIQ Toolkit ML Software Enablement environment.

Software Details

Features

  • Enhances the copyright protection to an ML model
  • Generates reports and artifacts and instructions needed to prove model ownership
  • No additional overhead to model performance or reduction in accuracy
  • Model Protection through Watermarking
    • Owning data is a differentiator to companies in the ML space
    • Collecting data and training a model is a substantial investment
    • Model watermarking protects your model intellectual property in the event it gets copied or stolen
  • Working Principles
    • The tool overlays a secret drawing on objects from a selected base class during training and labels these overlays as target class
    • When presented with base-class objects overlaid with the secret drawing during runtime the model will predict the target class instead of the base class hence identifying that the original model was copied

Supported Devices

  • i.MX8: i.MX 8 Family – Arm® Cortex®-A53, Cortex-A72, Virtualization, Vision, 3D Graphics, 4K Video
  • i.MX8M: i.MX 8M Family - Arm® Cortex®-A53, Cortex-M4, Audio, Voice, Video
  • i.MX8MMINI: i.MX 8M Mini - Arm® Cortex®-A53, Cortex-M4, Audio, Voice, Video
  • i.MX8MNANO: i.MX 8M Nano Family - Arm® Cortex®-A53, Cortex-M7
  • IMX8MPLUS: i.MX 8M Plus – Arm® Cortex®-A53, Machine Learning, Vision, Multimedia and Industrial IoT
  • i.MX8X: i.MX 8X Family – Arm® Cortex®-A35, 3D Graphics, 4K Video, DSP, Error Correcting Code on DDR
  • i.MX93: i.MX 93 Applications Processor Family – Arm® Cortex®-A55, ML Acceleration, Power Efficient MPU
  • iMX95: i.MX 95 Applications Processor Family: High-Performance, Safety Enabled Platform with eIQ® Neutron NPU
  • i.MX-RT500: i.MX RT500 Crossover MCU with Arm® Cortex®-M33, DSP and GPU Cores
  • i.MX-RT600: i.MX RT600 Crossover MCU with Arm® Cortex®-M33 and DSP Cores
  • i.MX-RT1050: i.MX RT1050 Crossover MCU with Arm® Cortex®-M7 Core
  • i.MX-RT1060: i.MX RT1060: Crossover MCU with Arm® Cortex®-M7
  • i.MX-RT1064: i.MX RT1064: Crossover MCU with Arm® Cortex®-M7
  • i.MX-RT1160: i.MX RT1160 Crossover MCU Dual-Core Arm® Cortex®-M7 and Cortex-M4
  • i.MX-RT1170: i.MX RT1170: 1 GHz Crossover MCU with Arm® Cortex® Cores

Downloads

Filters

1 download

Note: For better experience, software downloads are recommended on desktop.

Documentation

Quick reference to our documentation types.

Filters

1 documents

Training

Filters

1 trainings

Support

What do you need help with?

  • Narrow your search
Search