Black Box Membership Inference Against Object Detection Models

Abstract

For an unknown model use shadow models to create classifiers that will determine if data is training data or not. Evaluate unknown model results using classifiers. Uses Model's full prediction array.

Open PDF

Document Details

Document Type
Technical Report
Publication Date
Feb 10, 2021
Accession Number
AD1122793

Entities

People

  • Matthew Churilla

Organizations

  • Carnegie Mellon University

Tags

Communities of Interest

  • Autonomy

DTIC Thesaurus Topics

  • Boundaries
  • Classification
  • Computing-Related Activities
  • Contracts
  • Copyrights
  • Department Of Defense
  • Detection
  • Detectors
  • Engineering
  • Governments
  • Guarantees
  • Intellectual Property
  • Law
  • Learning
  • Machine Learning
  • Materials
  • Measurement
  • Patents
  • Software Development
  • Training
  • Universities

Fields of Study

  • Computer science

Readers

  • Computational Modeling and Simulation
  • Computer Vision.

Technology Areas

  • AI & ML
  • AI & ML - Bayesian Inference
  • AI & ML - Neural Networks