Metraipolis › archive › dc_efbadadf9e9d
Transformer-based Neural Beamforming for Real-Time Speech Enhancement on Smart Low-Power Hearable Devices — arXiv:2609.33755
- Source
- arXiv (en)
- Original
- https://arxiv.org/abs/2609.33755
- License (as archived)
- CC0-1.0
- Archived
- 2026-10-07T21:01:43.627Z
- SHA-256 of the archived text
dee64b1dc391b29fc3cbd4f69e68d27390427d3d22cc3342cea39e309f794037- Stable URL of this page
https://metraipolis.com/documento/dc_efbadadf9e9d
How to cite
Transformer-based Neural Beamforming for Real-Time Speech Enhancement on Smart Low-Power Hearable Devices — arXiv:2609.33755 — arXiv (en), https://arxiv.org/abs/2609.33755 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_efbadadf9e9d (Metraipolis archive dc_efbadadf9e9d, sha256 dee64b1dc391b29fc3cbd4f69e68d27390427d3d22cc3342cea39e309f794037)
BibTeX
@misc{dcefbadadf9e9d,
title = {Transformer-based Neural Beamforming for Real-Time Speech Enhancement on Smart Low-Power Hearable Devices — arXiv:2609.33755},
howpublished = {arXiv (en)},
url = {https://arxiv.org/abs/2609.33755},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_efbadadf9e9d; sha256 dee64b1dc391b29fc3cbd4f69e68d27390427d3d22cc3342cea39e309f794037},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.33755. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
Accurate, efficient, and low-latency spatial beamforming is a key component in emerging smart hearable devices, enhancing speech while suppressing noise and interference. However, handling multiple input sources under strict real-time constraints poses significant challenges for the low-power, resource-constrained microcontroller units (MCUs) used in hearables. We present an optimized methodology for the real-time execution of a neural-network-based minimum variance distortionless response (MVDR) beamformer on MCUs. Using six microphones and a three-stage mixed-precision scheme (float32 MVDR, int8 CNN, float16 Transformer), the pipeline pairs a CNN that estimates the MVDR weights with a lightweight Transformer that applies a per-frame correction. By time-slicing weight estimation with beamforming, it achieves a 15~ms per-frame latency while refreshing a complete set of CNN-derived weights every 564~ms. The deployed mixed-precision pipeline attains a short-time objective intelligibility (STOI) of 97.65\%, a scale-invariant signal-to-noise ratio (SI-SNR) of 20.26~dB, and a wideband PESQ of 3.676 at an average power of 45.9~mW. A speech activity detection (SAD) module (98.5\% accuracy, 0.62~mJ per inference) bypasses the pipeline during silence; under realistic deployment conditions, the system exceeds the 16~h all-day target on a 100~mAh battery, with an estimated lifetime of up to $\sim$20~h. To our knowledge, this is the first real-time multi-channel Transformer-based neural beamforming pipeline deployed on an MCU-class device. Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-09-29 in cs.AI). Categorie arXiv: Sound (cs.SD) ; Artificial Intelligence (cs.AI); Audio and Speech Processing (eess.AS). Pagina del lavoro: https://arxiv.org/abs/2609.33755 — 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 2026-10-07T20:29:02.730Z dalla pagina https://arxiv.org/catchup/cs.AI/2026-09-29?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).