Metraipolis › archive › dc_ca34aeeae916

BiTSE: Binaural Target Speaker Extraction in Noisy Multi-Talker Environments for AR Glass Arrays — arXiv:2608.10106

Source: arXiv (en)
Original: https://arxiv.org/abs/2608.10106
License (as archived): CC0-1.0
Archived: 2026-10-19T06:48:43.519Z
SHA-256 of the archived text: 505ffcf8c4b44c8ac6d7c4994624381519bb64be8da9ab4830d0897b14e69f17
Stable URL of this page: https://metraipolis.com/documento/dc_ca34aeeae916

How to cite

BiTSE: Binaural Target Speaker Extraction in Noisy Multi-Talker Environments for AR Glass Arrays — arXiv:2608.10106 — arXiv (en), https://arxiv.org/abs/2608.10106 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_ca34aeeae916 (Metraipolis archive dc_ca34aeeae916, sha256 505ffcf8c4b44c8ac6d7c4994624381519bb64be8da9ab4830d0897b14e69f17)

BibTeX

@misc{dcca34aeeae916,
  title = {BiTSE: Binaural Target Speaker Extraction in Noisy Multi-Talker Environments for AR Glass Arrays — arXiv:2608.10106},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2608.10106},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_ca34aeeae916; sha256 505ffcf8c4b44c8ac6d7c4994624381519bb64be8da9ab4830d0897b14e69f17},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2608.10106. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Isolating a desired speech signal in noisy multi-talker conversational scenarios is a key requirement for augmented reality (AR) wearable microphone array systems. In this work, a binaural target speaker extraction (TSE) framework, termed BiTSE, is proposed. It leverages both spatial and temporal cues, specifically the direction-of-arrival (DoA) of the target speaker and corresponding voice activity information, to guide the extraction process. Built upon a binaural signal denoising architecture, our model integrates three key enhancements: (i) a DoA-aware attention mechanism using cyclic positional embeddings, (ii) a timestamp-based masking strategy that utilizes speaker activity to suppress non-target segments, and (iii) a novel two-stage loss optimization strategy that first trains the model for robust denoising and then fine-tunes it to improve perceptual quality. Evaluations on the SPeech Enhancement for Augmented Reality (SPEAR) challenge dataset demonstrate that the proposed BiTSE consistently improves upon conventional approaches, leading to enhanced signal fidelity and perceptual quality.

Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-08-12 in eess.AS).
Categorie arXiv: Audio and Speech Processing (eess.AS).
Pagina del lavoro: https://arxiv.org/abs/2608.10106 — autori e testo completo sono lì (qui omessi i nomi: nessun dato personale nel deposito).
Licenza: titolo e abstract sono metadati arXiv distribuiti in CC0 1.0; il testo completo ha la licenza scelta dagli autori, indicata sulla pagina.
Letto il 2026-10-19T06:44:28.575Z dalla pagina https://arxiv.org/catchup/eess.AS/2026-08-12?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).