skip to main content


Search for: All records

Creators/Authors contains: "Zhao, T"

Note: When clicking on a Digital Object Identifier (DOI) number, you will be taken to an external site maintained by the publisher. Some full text articles may not yet be available without a charge during the embargo (administrative interval).
What is a DOI Number?

Some links on this page may take you to non-federal websites. Their policies may differ from this site.

  1. Free, publicly-accessible full text available July 1, 2024
  2. Hara, T. ; Yamaguchi, H. (Ed.)
    Prevalent wearables (e.g., smartwatches and activity trackers) demand high secure measures to protect users' private information, such as personal contacts, bank accounts, etc. While existing two-factor authentication methods can enhance traditional user authentication, they are not convenient as they require participations from users. Recently, manufacturing imperfections in hardware devices (e.g., accelerometers and WiFi interface) have been utilized for low-effort two-factor authentications. However, these methods rely on fixed device credentials that would require users to replace their devices once the device credentials are stolen. In this work, we develop a novel device authentication system, WatchID, that can identify a user's wearable using its vibration-based device credentials. Our system exploits readily available vibration motors and accelerometers in wearables to establish a vibration communication channel to capture wearables' unique vibration characteristics. Compared to existing methods, our vibration-based device credentials are reprogrammable and easy to use. We develop a series of data processing methods to mitigate the impact of noises and body movements. A lightweight convolutional neural network is developed for feature extraction and device authentication. Extensive experimental results using five smartwatches show that WatchID can achieve an average precision and recall of 98% and 94% respectively in various attacking scenarios. 
    more » « less
  3. The problem of continuous inverse optimal control (over finite time horizon) is to learn the unknown cost function over the sequence of continuous control variables from expert demonstrations. In this article, we study this fundamental problem in the framework of energy-based model, where the observed expert trajectories are assumed to be random samples from a probability density function defined as the exponential of the negative cost function up to a normalizing constant. The parameters of the cost function are learned by maximum likelihood via an “analysis by synthesis” scheme, which iterates (1) synthesis step: sample the synthesized trajectories from the current probability density using the Langevin dynamics via back-propagation through time, and (2) analysis step: update the model parameters based on the statistical difference between the synthesized trajectories and the observed trajectories. Given the fact that an efficient optimization algorithm is usually available for an optimal control problem, we also consider a convenient approximation of the above learning method, where we replace the sampling in the synthesis step by optimization. Moreover, to make the sampling or optimization more efficient, we propose to train the energy-based model simultaneously with a top-down trajectory generator via cooperative learning, where the trajectory generator is used to fast initialize the synthesis step of the energy-based model. We demonstrate the proposed methods on autonomous driving tasks, and show that they can learn suitable cost functions for optimal control. 
    more » « less
  4. null (Ed.)