Metraipolis › archive › dc_19fb5f36fc5b
U-PAST: A Phase-Aware Audio Spectrogram Transformer-U-Net for Single-Channel Speech Enhancement — arXiv:2609.00431
- Source
- arXiv (en)
- Original
- https://arxiv.org/abs/2609.00431
- License (as archived)
- CC0-1.0
- Archived
- 2026-10-17T07:34:09.617Z
- SHA-256 of the archived text
d8724ad8c3abad9f435db69bd10f204cc49c7f0a99a226856fe3153b14500680- Stable URL of this page
https://metraipolis.com/documento/dc_19fb5f36fc5b
How to cite
U-PAST: A Phase-Aware Audio Spectrogram Transformer-U-Net for Single-Channel Speech Enhancement — arXiv:2609.00431 — arXiv (en), https://arxiv.org/abs/2609.00431 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_19fb5f36fc5b (Metraipolis archive dc_19fb5f36fc5b, sha256 d8724ad8c3abad9f435db69bd10f204cc49c7f0a99a226856fe3153b14500680)
BibTeX
@misc{dc19fb5f36fc5b,
title = {U-PAST: A Phase-Aware Audio Spectrogram Transformer-U-Net for Single-Channel Speech Enhancement — arXiv:2609.00431},
howpublished = {arXiv (en)},
url = {https://arxiv.org/abs/2609.00431},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_19fb5f36fc5b; sha256 d8724ad8c3abad9f435db69bd10f204cc49c7f0a99a226856fe3153b14500680},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.00431. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
Convolutional neural networks (CNNs), used widely and successfully in audio enhancement, capture long-range time-frequency dependencies only indirectly, through successive convolution and pooling. Here, we present U-PAST, a hybrid transformer-U-Net architecture that addresses this limitation through self-attention dependency-modeling in the complex spectrogram domain. U-PAST tokenizes a complex STFT representation, similarly to the magnitude spectrogram tokenization of the Audio Spectrogram Transformer (AST), applies a multi-layer transformer encoder, and reconstructs the enhanced complex spectrogram with a U-Net-style decoder. We evaluate four architectural variants with between 1.17M and 2.40M parameters on the DNS Challenge, VoiceBank-DEMAND, and LibriMix corpora under matched, acoustic mismatch, and two-dataset mismatch conditions. U-PAST attains the best SI-SDR of any evaluated model under acoustic mismatch and closely trails substantially larger convolutional and time-domain baselines by 0.26 dB to 0.63 dB SI-SDR under the remaining three conditions while achieving the strongest perceptual (DNSMOS) quality under dataset mismatch. The largest evaluated configuration, U-PAST-H (2.40M parameters), is consistently the strongest variant of the family, offering an attractive performance-to-cost trade-off at a small parameter footprint. Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-09-02 in eess.AS). Categorie arXiv: Audio and Speech Processing (eess.AS) ; Sound (cs.SD). Pagina del lavoro: https://arxiv.org/abs/2609.00431 — 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 2026-10-17T07:27:11.528Z dalla pagina https://arxiv.org/catchup/eess.AS/2026-09-02?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).