Search arXiv⌕ Search
Home/Archive/arXiv

arXiv · 2509.02471

ESTM: An Enhanced Dual-Branch Spectral-Temporal Mamba for Anomalous Sound Detection

Chengyuan Ma·Peng Jia·Hongyue Guo·Wenming Yang
View original source ↗Read source document ↗Export metadata ↓

Abstract

The core challenge in industrial equipment anoma lous sound detection (ASD) lies in modeling the time-frequency coupling characteristics of acoustic features. Existing modeling methods are limited by local receptive fields, making it difficult to capture long-range temporal patterns and cross-band dynamic coupling effects in machine acoustic features. In this paper, we propose a novel framework, ESTM, which is based on a dual-path Mamba architecture with time-frequency decoupled modeling and utilizes Selective State-Space Models (SSM) for long-range sequence modeling. ESTM extracts rich feature representations from different time segments and frequency bands by fusing enhanced Mel spectrograms and raw audio features, while further improving sensitivity to anomalous patterns through the TriStat-Gating (TSG) module. Our experiments demonstrate that ESTM improves anomalous detection performance on the DCASE 2020 Task 2 dataset, further validating the effectiveness of the proposed method.

Explore related subjects

cs.SDcs.LG

Keep this discovery

Explore connections, maps & timelines

BibTeXRIS

Chengyuan Ma, Peng Jia, Hongyue Guo, Wenming Yang. 2025-09-02. ESTM: An Enhanced Dual-Branch Spectral-Temporal Mamba for Anomalous Sound Detection. https://arxiv.org/abs/2509.02471

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

KEEP EXPLORING

Related papers

arXiv2026

Streaming Generation for Music Accompaniment

Music generation models can produce high-fidelity coherent accompaniment given complete audio input, but are limited to editing and loop-based workflows. We study real-time audio-to-audio accompaniment: as a model hears an input audio stream (e.g., a singer singing), it has to also simultaneously generate in real-time a coherent accompanying stream (e.g., a guitar accompaniment). In this work, we propose a model design considering inevitable system delays in practical deployment with two design variables: future visibility $t_f$, the offset between the output playback time and the latest input time used for conditioning, and output chunk duration $k$, the number of frames emitted per call. We train Transformer decoders across a grid of $(t_f,k)$ and show two consistent trade-offs: increasing effective $t_f$ improves coherence by reducing the recency gap, but requires faster inference to stay within the latency budget; increasing $k$ improves throughput but results in degraded accompaniment due to a reduced update rate. Finally, we observe that naive maximum-likelihood streaming training is insufficient for coherent accompaniment where future context is not available, motivating advanced anticipatory and agentic objectives for live jamming.

cs.SD↗
arXiv2026

Decoding Order Matters in Autoregressive Speech Synthesis

Autoregressive speech synthesis often adopts a left-to-right order, yet generation order is a modelling choice. We investigate decoding order through masked diffusion framework, which progressively unmasks positions and allows arbitrary decoding orders during training and inference. By interpolating between identity and random permutations, we show that randomness in decoding order affects speech quality. We further compare fixed strategies, such as \texttt{l2r} and \texttt{r2l} with adaptive ones, such as Top-$K$, finding that fixed-order decoding, including the dominating left-to-right approach, is suboptimal, while adaptive decoding yields better performance. Finally, since masked diffusion requires discrete inputs, we quantise acoustic representations and find that even 1-bit quantisation can support reasonably high-quality speech.

cs.SD↗
arXiv2026

An Interpretable, Controllable Time-Varying IIR Denoiser for On-Device Assistive Hearing

We present TVBC (Time-Varying Biquad Cascade), an interpretable, low-latency speech enhancement model for real-time, on-device assistive hearing. A lightweight neural controller predicts, in real time, the coefficients of a differentiable cascade of 35 second-order IIR filters (biquads), so the model tracks non-stationary noise while keeping a fully interpretable processing chain: every spectral modification is an explicit, adjustable equalizer curve rather than an opaque `black-box' transform. Because the biquad cascade carries the signal processing, the controller can be made very small, driving the cascade with only 24k parameters at a 10.7ms algorithmic latency, within hearing-aid budgets, and running entirely on-device so that audio never leaves the device. We also expose the suppression-versus-preservation trade-off as an explicit control: it can be set during training through the loss weighting, and adjusted at inference, with no retraining, by mixing the noisy input with the denoised output. On hearing-aid metrics (HASPI/HASQI) the 24k model stays within about 0.02 of DFNet3 (2.3M parameters, almost two orders of magnitude larger) while using about 29X fewer multiply-accumulates, although larger black-box models still lead on reference metrics such as PESQ. We present TVBC as a proof of concept for a compact, interpretable, and controllable denoiser for on-device assistive hearing.

cs.SD↗