Searcharxiv⌕ Search
Home/Archive/arXiv

arXiv · 2402.06411

Exploiting spatial diversity for increasing the robustness of sound source localization systems against reverberation

Guillermo Garcia-Barrios·Eduardo Latorre Iglesias·Juana M. Gutierrez-Arriola·Ruben Fraile·Nicolas Saenz-Lechon·Victor Jose Osma-Ruiz
View original source ↗Read source document ↗Export metadata ↓

Abstract

Acoustic reverberation is one of the most relevant factors that hampers the localization of a sound source inside a room. To date, several approaches have been proposed to deal with it, but have not always been evaluated under realistic conditions. This paper proposes exploiting spatial diversity as an alternative approach to achieve robustness against reverberation. The theoretical arguments supporting this approach are first presented and later confirmed by means of simulation results and real measurements. Simulations are run for reverberation times up to 2 s, thus providing results with a wider range of validity than in other previous research works. It is concluded that the use of systems consisting of several, sufficiently separated, small arrays leads to the best results in reverberant environments. Some recommendations are given regarding the choice of the array sizes, the separation among them, and the way to combine SRP-PHAT maps obtained from diverse arrays.

Explore related subjects

cs.SDeess.ASeess.SP

Keep this discovery

BibTeXRIS

Guillermo Garcia-Barrios, Eduardo Latorre Iglesias, Juana M. Gutierrez-Arriola, Ruben Fraile, Nicolas Saenz-Lechon, Victor Jose Osma-Ruiz. 2024-02-09. Exploiting spatial diversity for increasing the robustness of sound source localization systems against reverberation. https://doi.org/10.1016/j.apacoust.2022.109138

Cite the original work for its findings. Save a collection to share your selection of sources.

KEEP EXPLORING

Related papers

arXiv2026

Xiaomi-CocktailASR-1 Technical Report

Recently, large language model (LLM) based ASR models have achieved significant progress, yet they generally lack support for multi-speaker scenarios, where the cocktail party problem remains a critical bottleneck for further advancing ASR. Existing TS-ASR methods, including end-to-end architectures with speaker embeddings and latest LLM-based explorations suffer from degraded single-speaker performance and the inability to reject when the target speaker is absent. In this paper, we propose Xiaomi-CocktailASR-1, an LLM-based end-to-end TS-ASR architecture. By utilizing reference speech as voiceprint prompts, it directly transcribes the target speaker's speech without requiring speech separation. Xiaomi-CocktailASR-1 maintains competitive performance in single-speaker scenarios, comparable to mainstream ASR models. It also features a negative sample rejection capability, outputting empty text when the target speaker is absent from the mixed speech. Additionally, Xiaomi-CocktailASR-1 supports a Chain-of-Thought (CoT) reasoning mode to provide explicit reasoning steps. Extensive experiments on various synthetic and real-world multispeaker benchmarks demonstrate that Xiaomi-CocktailASR-1 achieves state-of-the-art performance, effectively addressing the cocktail party problem through a unified architecture that balances multispeaker and single-speaker recognition accuracy, along with rejection capability.

cs.SD↗
arXiv2026

EConv-TasNet: Efficient Conv-TasNet for Effective Speech Separation

Conv-TasNet has served as a strong baseline for time-domain speech separation, and many studies have extended it with advanced architectures such as dual-path networks, U-Nets, and attention mechanisms. However, these methods often introduce high computational cost and complexity, limiting their deployment in resource-constrained scenarios. To address this issue, we propose eConv-TasNet, an efficient variant of Conv-TasNet that improves both effectiveness and efficiency without relying on resource-intensive modules. The proposed model consists of a group-wise early-splitting (GES) module and a multi-group feature aggregation (MGFA) module. GES generates discriminative speaker embeddings at intermediate stages, while MGFA progressively aggregates these group-level representations for refined mask estimation. Experimental results show that eConv-TasNet reduces model size by 22.4%, accelerates inference by 18.9%, and improves SI-SNRi by 14.0%-28.0% across three public benchmarks. Moreover, it achieves competitive performance compared with state-of-the-art methods while requiring significantly fewer parameters and lower inference cost. These results demonstrate a favorable efficiency-effectiveness trade-off for edge deployment.

cs.SD↗