Search

Sound event localization and detection (SELD) systems using audio recordings from a microphone array rely on spatial cues for determining the location of sound events. As a consequence, the localization performance of such systems is to a large extent determined by the quality of the audio features that are used as inputs to the system. We propose a new feature, based on neural generalized cross-correlations with phase-transform (NGCC-PHAT), that learns audio representations suitable for localization. Using permutation invariant training for the time-difference of arrival (TDOA) estimation problem enables NGCC-PHAT to learn TDOA features for multiple overlapping sound events. These features can be used as a drop-in replacement for GCC-PHAT inputs to a SELD-network. We test our method on the STARSS23 dataset and demonstrate improved localization performance compared to using standard GCC-PHAT or SALSA-Lite input features., Comment: DCASE 2024

We present a novel approach to the 3D sound source localization task for distributed ad-hoc microphone arrays by formulating it as a set-to-set regression problem. By training a multi-modal masked autoencoder model that operates on audio recordings and microphone coordinates, we show that such a formulation allows for accurate localization of the sound source, by reconstructing coordinates masked in the input. Our approach is flexible in the sense that a single model can be used with an arbitrary number of microphones, even when a subset of audio recordings and microphone coordinates are missing. We test our method on simulated and real-world recordings of music and speech in indoor environments, and demonstrate competitive performance compared to both classical and other learning based localization methods., Comment: IPIN 2024

We present a unique comparative analysis, and evaluation of vision, radio, and audio based localization algorithms. We create the first baseline for the aforementioned sensors using the recently published Lund University Vision, Radio, and Audio (LuViRA) dataset, where all the sensors are synchronized and measured in the same environment. Some of the challenges of using each specific sensor for indoor localization tasks are highlighted. Each sensor is paired with a current state-of-the-art localization algorithm and evaluated for different aspects: localization accuracy, reliability and sensitivity to environment changes, calibration requirements, and potential system complexity. Specifically, the evaluation covers the ORB-SLAM3 algorithm for vision-based localization with an RGB-D camera, a machine-learning algorithm for radio-based localization with massive MIMO technology, and the SFS2 algorithm for audio-based localization with distributed microphones. The results can serve as a guideline and basis for further development of robust and high-precision multi-sensory localization systems, e.g., through sensor fusion, context, and environment-aware adaptation., Comment: 10 pages, 11 figures

Nowadays, state-of-the-art learning-to-rank methods are based on gradient-boosted decision trees (GBDT). The most well-known algorithm is LambdaMART which was proposed more than a decade ago. Recently, several other GBDT-based ranking algorithms were proposed. In this paper, we thoroughly analyze these methods in a unified setup. In particular, we address the following questions. Is direct optimization of a smoothed ranking loss preferable over optimizing a convex surrogate? How to properly construct and smooth surrogate ranking losses? To address these questions, we compare LambdaMART with YetiRank and StochasticRank methods and their modifications. We also propose a simple improvement of the YetiRank approach that allows for optimizing specific ranking loss functions. As a result, we gain insights into learning-to-rank techniques and obtain a new state-of-the-art algorithm.

Mobile robots have the capability to work in real-time autonomously. Autonomous behavior is strictly dependent on knowing the position of the mobile robot. The positioning of a mobile robot in an indoor area is a difficult task for only one sensor information is used. We proposed a system and method to locate the mobile robot via fusing signals from WIFI and odometer data via particle filter. In this study, the Particle filter is a well-known filter that is used for indoor positioning of mobile robots. The proposed system includes two parts that are RFKON system and evarobot for data collection and experiments. The Received Signal Strength (RSS) measurements of the WiFi access points that are located in any environment are used to locate a stationary mobile robot in one floor area via SIS Particle Filter. RSS measurements from the RFKON database are used and the average location error is 0.7606 and 0.1495 m for 300 and 1000 particles respectively., Comment: 12 pages, 15 figures, 1 table, This work is supported by The Scientific and Technological Research Council of Turkey (TUBITAK) under "RF based indoor positioning system (RFKON)" grant number 1130024

Graph Convolutional Networks (GCNs) have gained significant developments in representation learning on graphs. However, current GCNs suffer from two common challenges: 1) GCNs are only effective with shallow structures; stacking multiple GCN layers will lead to over-smoothing. 2) GCNs do not scale well with large, dense graphs due to the recursive neighborhood expansion. We generalize the propagation strategies of current GCNs as a \emph{"Sink$\to$Source"} mode, which seems to be an underlying cause of the two challenges. To address these issues intrinsically, in this paper, we study the information propagation mechanism in a \emph{"Source$\to$Sink"} mode. We introduce a new concept "information flow path" that explicitly defines where information originates and how it diffuses. Then a novel framework, namely Flow Graph Network (FlowGN), is proposed to learn node representations. FlowGN is computationally efficient and flexible in propagation strategies. Moreover, FlowGN decouples the layer structure from the information propagation process, removing the interior constraint of applying deep structures in traditional GCNs. Further experiments on public datasets demonstrate the superiority of FlowGN against state-of-the-art GCNs.

In this paper we present CatBoost, a new open-sourced gradient boosting library that successfully handles categorical features and outperforms existing publicly available implementations of gradient boosting in terms of quality on a set of popular publicly available datasets. The library has a GPU implementation of learning algorithm and a CPU implementation of scoring algorithm, which are significantly faster than other gradient boosting libraries on ensembles of similar sizes.

We present a direct, geometric derivation of the generalized Smarr formula for the stationary axially symmetric black holes with nonlinear electromagnetic fields. The additional term is proven to be proportional to the integral of the trace of the electromagnetic energy-momentum tensor and can be written as a product of two conjugate variables. From the novel relation we can deduce all previously proposed forms of the generalized Smarr formula, which were derived only for the spherically symmetric black holes, and provide the lowest order quantum correction to the classical relation from the Euler-Heisenberg Lagrangian., Comment: 20 pages, v3: slightly revised, published version

This paper presents the key algorithmic techniques behind CatBoost, a new gradient boosting toolkit. Their combination leads to CatBoost outperforming other publicly available boosting implementations in terms of quality on a variety of datasets. Two critical algorithmic advances introduced in CatBoost are the implementation of ordered boosting, a permutation-driven alternative to the classic algorithm, and an innovative algorithm for processing categorical features. Both techniques were created to fight a prediction shift caused by a special kind of target leakage present in all currently existing implementations of gradient boosting algorithms. In this paper, we provide a detailed analysis of this problem and demonstrate that proposed algorithms solve it effectively, leading to excellent empirical results.

We extend the classical results on the symmetry inheritance of the canonical electromagnetic fields, described by the Maxwell's Lagrangian, to a much wider class of models, which include those of the Born-Infeld, power Maxwell and the Euler-Heisenberg type. Symmetry inheriting fields allow the introduction of electromagnetic scalar potentials and these are proven to be constant on the Killing horizons. Finally, using the relations obtained along the analysis, we generalize and simplify the recent proof for the symmetry inheritance of the 3-dimensional case, as well as give the first constraint for the higher dimensional electromagnetic fields., Comment: 6 pages; slightly revised, published version (two statements removed from the Theorem II.1, equations (25) and (26) corrected, several comments added/revised)

We present the results of the first experimental study of ionization yield of electron recoils with energies below 100 keV produced in liquid xenon by the isotopes: 37Ar, 83mKr, 241Am, 129Xe, 131Xe. It is confirmed by a direct measurement with 37Ar isotope (2.82 keV) that the ionization yield is growing up with the energy decrease in the energy range below ~ 10 keV accordingly to the NEST predictions. Decay time of scintillation at 2.82 keV is measured to be 25 +/- 3 ns at the electric field of 3.75 kV/cm., Comment: 16 pages, 8 figures

We propose to detect and to study neutrino neutral current coherent scattering off atomic nuclei with a two-phase emission detector using liquid xenon as a working medium. Expected signals and backgrounds are calculated for two possible experimental sites: Kalinin Nuclear Power Plant in the Russian Federation and Spallation Neutron Source at the Oak Ridge National Laboratory in the USA. Both sites have advantages as well as limitations. However the experiment looks feasible at either location. Preliminary design of the detector and supporting R&D program are discussed., Comment: 16 pages, 10 figures, RED collaboration