skip to main content


Title: Expedited Multi-Target Search with Guaranteed Performance via Multi-fidelity Gaussian Processes
We consider a scenario in which an autonomous vehicle equipped with a downward-facing camera operates in a 3D environment and is tasked with searching for an unknown number of stationary targets on the 2D floor of the environment. The key challenge is to minimize the search time while ensuring a high detection accuracy. We model the sensing field using a multi-fidelity Gaussian process that systematically describes the sensing information available at different altitudes from the floor. Based on the sensing model, we design a novel algorithm called Expedited Multi-Target Search (EMTS) that (i) addresses the coverage-accuracy trade-off: sampling at locations farther from the floor provides a wider field of view but less accurate measurements, (ii) computes an occupancy map of the floor within a prescribed accuracy and quickly eliminates unoccupied regions from the search space, and (iii) travels efficiently to collect the required samples for target detection. We rigorously analyze the algorithm and establish formal guarantees on the target detection accuracy and the detection time. We illustrate the algorithm using a simulated multi-target search scenario.  more » « less
Award ID(s):
1734272
NSF-PAR ID:
10309128
Author(s) / Creator(s):
; ;
Date Published:
Journal Name:
2020 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS)
Format(s):
Medium: X
Sponsoring Org:
National Science Foundation
More Like this
  1. Sensor fusion approaches combine data from a suite of sensors into an integrated solution that represents the target environment more accurately than that produced by an individual sensor. Deep learning (DL) based approaches can address challenges with sensor fusion more accurately than classical approaches. However, the accuracy of the selected approach can change when sensors are modified, upgraded or swapped out within the system of sensors. Historically, this can require an expensive manual refactor of the sensor fusion solution.This paper develops 12 DL-based sensor fusion approaches and proposes a systematic and iterative methodology for selecting an optimal DL approach and hyperparameter settings simultaneously. The Gradient Descent Multi-Algorithm Grid Search (GD-MAGS) methodology is an iterative grid search technique enhanced by gradient descent predictions and expanded to exchange performance measure information across concurrently running DL-based approaches. Additionally, at each iteration, the worst two performing DL approaches are pruned to reduce the resource usage as computational expense increases from hyperparameter tuning. We evaluate this methodology using an open source, time-series aircraft data set trained on the aircraft’s altitude using multi-modal sensors that measure variables such as velocities, accelerations, pressures, temperatures, and aircraft orientation and position. We demonstrate the selection of an optimal DL model and an increase of 88% in model accuracy compared to the other 11 DL approaches analyzed. Verification of the model selected shows that it outperforms pruned models on data from other aircraft with the same system of sensors. 
    more » « less
  2. Presented at the Workshop on Heterogeneous Multi-Robot Task Allocation and Coordination. The authors recently developed a distributed algorithm to enable a team of homogeneous robots to search for and track an unknown and time-varying number of dynamic targets. This algorithm combined a distributed version of the PHD filter (for multi-target tracking) with Lloyd’s algorithm to drive the motion of the robots. In this paper we extend this previous work to allow a heterogeneous team of groundand aerial robots to perform the search and tracking tasks in a coordinated manner. Both types of robots are equipped with sensors that have a finite field of view and which may receive both false positive and false negative detections. Theaerial robots may vary the size of their sensor field of view (FoV) by changing elevation. This increase in the FoV coincides with a decrease in the accuracy and reliability of the sensor. The ground robots maintain the target tracking information while the aerial robots provide additional sensor coverage. We develop two new distributed algorithms to provide filter updates and to make control decisions in this heterogeneous team. Both algorithms only require robots to communicate with nearby robots and use minimal bandwidth.We demonstrate the efficacy of our approach through a series of simulated experiments which show that the heterogeneous teams are able to achieve more accurate tracking in less time than our previous work. 
    more » « less
  3. This paper compares different distributed control approaches that enable a team of robots search for and track an unknown number of targets. The robots are equipped with sensors which have limited field of view (FoV) and are required to explore the environment. The team uses a distributed formulation of the Probability Hypothesis Density (PHD) filter to estimate the number and the position of the targets. The resulting target estimate is used to select the future search locations for each robot. This paper compares Lloyd’s algorithm, a traditional method for distributed search, with two typical stochastic optimization methods, Particle Swarm Optimization (PSO) and Simulated Annealing (SA). PSO and SA are traditionally used to find a single global maximum, therefore this paper describes novel formulations of PSO and SA to solve the problem of multi-target tracking. These new methods more effectively trade off between exploration and exploitation. Simulations demonstrate that the use of these stochastic optimization techniques improves coverage of the search space and reduces the error in the target estimates compared to the baseline approach. 
    more » « less
  4. Distributed multi-target tracking is a canonical task for multi-robot systems, encompassing applications from environmental monitoring to disaster response to surveillance. In many situations, the distribution of unknown objects in a search area is irregular, with objects are likely to distribute in clusters instead of evenly distributed. In this paper, we develop a novel distributed multi-robot multi-target tracking algorithm for effectively tracking clustered targets from noisy measurements. Our algorithm contains two major components. Firstly, both the instantaneous and cumulative target density are estimated, providing the best guess of current target states and long-term coarse distribution of clusters, respectively. Secondly, the power diagram is implemented in Lloyd’s algorithm to optimize task space assignment for each robot to trade-off between tracking detected targets in clusters and searching for potential targets outside clusters. We demonstrate the efficacy of our proposed method and show that our method outperforms of other candidates in tracking accuracy through a set of simulations. 
    more » « less
  5. Abstract

    Current radio frequency (RF) classification techniques assume only one target in the field of view. Multi‐target recognition is challenging because conventional radar signal processing results in the superposition of target micro‐Doppler signatures, making it difficult to recognise multi‐target activity. This study proposes an angular subspace projection technique that generates multiple radar data cubes (RDC) conditioned on angle (RDC‐ω). This approach enables signal separation in the raw RDC, making possible the utilisation of deep neural networks taking the raw RF data as input or any other data representation in multi‐target scenarios. When targets are in closer proximity and cannot be separated by classical techniques, the proposed approach boosts the relative signal‐to‐noise ratio between targets, resulting in multi‐view spectrograms that boosts the classification accuracy when input to the proposed multi‐view DNN. Our results qualitatively and quantitatively characterise the similarity of multi‐view signatures to those acquired in a single‐target configuration. For a nine‐class activity recognition problem, 97.8% accuracy in a 3‐person scenario is achieved, while utilising DNN trained on single‐target data. We also present the results for two cases of close proximity (sign language recognition and side‐by‐side activities), where the proposed approach has boosted the performance.

     
    more » « less