Adversarial Plannning

  title={Adversarial Plannning},
  author={Valentin Vie and Ryan Sheatsley and Sophia Beyda and Sushrut Shringarputale and Kevin S. Chan and Trent Jaeger and Patrick Mcdaniel},
Planning algorithms are used in computational systems to direct autonomous behavior. In a canonical application for example, planning for autonomous vehicles is used to automate the static or continuous planning towards performance, resource management, or functional goals (e.g., arriving at the destination, managing fuel consumption). Existing planning algorithms assume non-adversarial settings; a least cost plan is developed based on available environmental information (i.e., the input… 



Course of Action Generation for Cyber Security Using Classical Planning

In this application, the generation of attack plans for a simple but realistic web-based document control system is demonstrated, with excellent performance compared to the prevailing state of the art in this area.

Adversarial examples in the physical world

It is found that a large fraction of adversarial examples are classified incorrectly even when perceived through the camera, which shows that even in physical world scenarios, machine learning systems are vulnerable to adversarialExamples.

A PCA-Based Model to Predict Adversarial Examples on Q-Learning of Path Finding

A probabilistic output model based on the influence factors and the corresponding weights to predict the adversarial examples of Q-learning in automatic path finding is proposed and can reach to 70% with the proper parameter setting.

Towards Evaluating the Robustness of Neural Networks

It is demonstrated that defensive distillation does not significantly increase the robustness of neural networks, and three new attack algorithms are introduced that are successful on both distilled and undistilled neural networks with 100% probability are introduced.

A complete navigation system for goal acquisition in unknown environments

  • A. StentzM. Hebert
  • Computer Science
    Proceedings 1995 IEEE/RSJ International Conference on Intelligent Robots and Systems. Human Robot Interaction and Cooperative Robots
  • 1995
This work has developed a complete system that integrates local and global navigation that was tested on a real robot and successfully drove it 1.4 kilometers to find a goal given no a priori map of the environment.

Security vulnerabilities of connected vehicle streams and their impact on cooperative driving

This study presents a first look at the effects of security attacks on the communication channel as well as sensor tampering of a connected vehicle stream equipped to achieve CACC, and shows that an insider attack can cause significant instability in the CACC vehicle stream.

Adversarial Machine Learning at Scale

This research applies adversarial training to ImageNet and finds that single-step attacks are the best for mounting black-box attacks, and resolution of a "label leaking" effect that causes adversarially trained models to perform better on adversarial examples than on clean examples.

Engineering Benchmarks for Planning: the Domains Used in the Deterministic Part of IPC-4

The article explains and discusses the five application domains and their adaptation to form the PDDL test suites used in IPC-4, and summarizes known theoretical results on structural properties of the domains, regarding their computational complexity and provable properties of their topology under the h+ function.

PDDL2.1: An Extension to PDDL for Expressing Temporal Planning Domains

The syntax of the language, PDDL2.1, is described, which has considerable modelling power -- exceeding the capabilities of current planning technology -- and presents a number of important challenges to the research community.

The Limitations of Deep Learning in Adversarial Settings

This work formalizes the space of adversaries against deep neural networks (DNNs) and introduces a novel class of algorithms to craft adversarial samples based on a precise understanding of the mapping between inputs and outputs of DNNs.