Inlier Maximization for Robust Direction-of-Arrival Estimation in Ad-hoc Microphone Networks

Gabrielle Flood, Kalle Åström, Magnus Oskarsson

Left: the microphone setup used in the experiments, Middle: detected medium sized quadcopter and fixed wing drone, Right: DOA estimates on one test sequence.

Abstract

We present a robust method for estimating the direction-of-arrival of a target, given time-difference-of-arrival measurements. The method is based on a theoretic framework for optimal inlier maximization, and we present efficient solvers that can be used in a polynomial time algorithm, with guaranteed global optimality. We especially investigate the problem where the data come from microphone recordings in an ad-hoc audio network. We show, on synthetic and real data, that our method is competitive in terms of both robustness and speed.

Acknowledgement

This work was supported by the Czech Science Foundation (GACR) JUNIOR STAR, Grant No. 22-23183M, by the strategic research project ELLIIT, and partially supported by the Wallenberg AI, Autonomous Systems and Software Program (WASP) funded by the Knut and Alice Wallenberg Foundation. The authors are grateful to Erik Tegler for previous collaborations and valuable discussions leading up to this paper.

BibTeX


      @inproceedings{flood2026inlier,
        title={Inlier Maximization for Robust Direction-of-Arrival Estimation in Ad-hoc Microphone Networks},
        author={Flood, Gabrielle and {\AA}str{\"o}m, Kalle and Oskarsson, Magnus},
        booktitle={34th European Signal Processing Conference (EUSIPCO)},
        year={2026}
      }