DCDemetris Coleman

Demetris Coleman

*This is an algorithmically generated author record based on publicly available information.

Current Organizations / Institutes

The organizations/institutes from the scholar's latest article

Previous Organizations / Institutes

The organizations / institutes from the scholar's other articles in the past ten years


Demetris Coleman has a total of 7 co-authors. Here are the top co-authors by number of shared publications:


Latest Publications

  • Open Access

    Adaptive exploration under localization uncertainty using multi-fidelity Gaussian processes

    by Demetris Coleman, Shaunak D. Bopardikar, Vaibhav Srivastava, Xiaobo Tan

    This paper considers a robot moving in a 3D environment that is tasked with estimating a quasi-stationary environmental field (e.g., temperature, concentration of a chemical pollutant, or distribution of light radiation density) in the presence of localization uncertainties, as is typical in underwater or other GPS-denied environments. Gaussian process regression has been widely adopted to model environmental fields. However, a drawback of Gaussian process regression is its difficulty in accounting for data with uncertain input. This work proposes a novel multi-fidelity Gaussian process-based regression approach to address the challenge by splitting the data collected by the robot into different datasets corresponding to the amount of input (localization) uncertainty. Furthermore, a sampling-based trajectory planning algorithm is proposed for adaptive robot exploration that optimizes a field-reconstruction objective function while accommodating resource constraints. The proposed approach is experimentally evaluated using a miniature gliding robotic fish that measures light intensity in a large indoor tank. The adaptive exploration algorithm is tested using both a multi-fidelity Gaussian process model and a baseline single-fidelity model. Two objective functions, based on the information gain and an ergodic metric, respectively, are adopted in the evaluation. The experiments show that, for both objective functions, using multi-fidelity Gaussian process reduces the weighted mean squared error between the model prediction and the ground-truth field compared to using the baseline single-fidelity model that ignores localization uncertainty. Accompanying code available at Coleman (Adaptive exploration under localization uncertainty using multi-fidelity Gaussian processes, 2025, https://github.com/colem404/Adaptive-Exploration-Under-Localization-Uncertainty-Using-Multi-fidelity-Gaussian-Processes/tree/main ).


  • Exploration of Unknown Scalar Fields with Multifidelity Gaussian Processes Under Localization Uncertainty

    by Demetris Coleman, Shaunak D. Bopardikar, Vaibhav Srivastava, Xiaobo Tan

    Autonomous marine vehicles are deployed in oceans and lakes to collect spatio-temporal data. GPS is often used for localization, but is inaccessible underwater. Poor localization underwater makes it difficult to pinpoint where data are collected, to accurately map, or to autonomously explore the ocean and other aquatic environments. This paper proposes the use of multifidelity Gaussian process regression to incorporate data associated with uncertain locations. With the proposed approach, an adaptive sampling algorithm is developed for exploration and mapping of unknown scalar fields. The reconstruction performance based on the multifidelity model is compared to that based on a single-fidelity Gaussian process model that only uses data with known locations, and to that based on a single-fidelity Gaussian process model that ignores the localization error. Numerical results show that the proposed multifidelity approach outperforms both single-fidelity approaches in terms of the reconstruction accuracy.


  • Incorporating Observability via Control Barrier Functions with Application to Range-based Target Tracking

    by Demetris Coleman, Shaunak D. Bopardikar, Xiaobo Tan

    In nonlinear systems, the control input often directly impacts observability of the system. In this paper, we investigate the use of control barrier functions (CBFs) for enforcing observability of a mobile robot in target tracking, when only the distance to the target is measured. The problem is motivated by practical applications for autonomous robots when operating in GPS-denied environments. To address the tradeoffs between localization accuracy and tracking performance, a tracking controller is augmented by a control barrier function based on an observability metric. Two examples are used to show the efficacy of the approach, one with unicycle dynamics on a plane, and the other based on gliding robotic fish with complex 3D dynamics. The approach taken in this work is compared to a model predictive controller that optimizes a joint cost function of tracking error and observability metric. While both approaches are shown to maintain observability and enable tracking, the CBF-based approach is shown to have several advantages


  • Open Access

    Towards Safer Self-Driving Through Great PAIN (Physically Adversarial Intelligent Networks)

    by Piyush Gupta, Demetris Coleman, Joshua E. Siegel

    Automated vehicles' neural networks suffer from overfit, poor generalizability, and untrained edge cases due to limited data availability. Researchers synthesize randomized edge-case scenarios to assist in the training process, though simulation introduces potential for overfit to latent rules and features. Automating worst-case scenario generation could yield informative data for improving self driving. To this end, we introduce a "Physically Adversarial Intelligent Network" (PAIN), wherein self-driving vehicles interact aggressively in the CARLA simulation environment. We train two agents, a protagonist and an adversary, using dueling double deep Q networks (DDDQNs) with prioritized experience replay. The coupled networks alternately seek-to-collide and to avoid collisions such that the "defensive" avoidance algorithm increases the mean-time-to-failure and distance traveled under non-hostile operating conditions. The trained protagonist becomes more resilient to environmental uncertainty and less prone to corner case failures resulting in collisions than the agent trained without an adversary.