Peer Revenue Publish or profit

Papers

Fresh from arXiv, typeset properly. Highlighted words trade on the market.

Electrical Engineering and Systems Science Audio and Speech Processing

eess.AS Watch Audio and Speech Processing

Tuesday 29 September

Audio and Speech Processing · Cross-listed from cs.SD

Reconstructing the Vocal Tract with Differentiable Acoustic Simulation

Eric Ming Chen, Jin Woo Lee and Vincent Sitzmann

The vocal tract is the region of the human body responsible for filtering one's voice to create speech. In this paper, we present a differentiable and GPU accelerated acoustic simulator for the vocal tract. The differentiable simulator synthesizes speech by propagating sound along an acoustic tube model of the vocal tract, and via its gradients, can solve the inverse problem: reconstructing the shape of the vocal tract solely from the sound it produces. Although the inverse mapping between geometry and sound is notoriously non-convex, we discover that gradient descent succeeds with three technical contributions: (1) we design a frequency domain formulation of the vocal tract's fluid dynamics that is 70x more GPU parallelizable than finite differences in time, (2) we integrate a differentiable model for turbulence to synthesize consonants, and (3) similar to prior work in implicit neural representations (INRs) and neural fields, we find that parameterizing the geometry with a neural network accelerates convergence and escapes local minima that trap discrete representations. Because the simulator is differentiable, it is readily integrated with other deep learning pipelines to enable novel linguistics and medical imaging applications. (1) We demonstrate self-supervised autoencoding of vocal tract shapes across 11 languages, and (2) we couple our simulator with a generative model of MRI (magnetic resonance imaging) images to reconstruct one's moving vocal tract from only their speech without paired data.
Read the full abstract

Audio and Speech Processing

Perception-Inspired Bayesian Causal Fusion for Audiovisual Source Localization

Kyung Yun Lee, Sungnyun Kim, Sebastian J. Schlecht et al. (+2 more)

Multimodal fusion promises more accurate perception but only when the modalities share a common cause. When they do not, the second modality carries no information about the target, and fusing it can only corrupt the estimate. We cast this whether-to-fuse decision as Bayesian causal inference, following the optimal-observer model of human multisensory perception, and implement it as a plug-and-play layer on top of frozen audio and visual models for sound event localization and detection. The model infers a common-cause posterior over visible candidates, then gates precision-weighted fusion accordingly. Fusing unconditionally more than doubles the direction error, whereas the causal gate improves on-screen localization while limiting off-screen degradation, without any joint network retraining.
Read the full abstract

All authors Kyung Yun Lee, Sungnyun Kim, Sebastian J. Schlecht, Tae-Hyun Oh, Vesa Välimäki

Monday 28 September

Audio and Speech Processing · Cross-listed from cs.SD

When Does a Spoken Agent Have Enough Evidence to Act? The PACT-SLM Contract Test

Mengzhe Geng

Streaming spoken agents may produce the correct final action after acting too early. Final-turn scores do not reveal whether each observed speech prefix supports an exposed action. We introduce the Partial Speech Action Contract for Turn Taking in Speech Language Models (PACT-SLM), a controlled test that assigns the first valid action time and evaluates both action identity and timing. In the primary test, 80 paired contrast groups from four held-out semantic families yield 1,600 prefix predictions across clean and 15 dB noise renderings. Using source-utterance semantic targets rather than counterbalanced branch codes, WavLM Base Plus reaches 26.03% pooled post-onset semantic-label accuracy (95% group-bootstrap interval [22.14%, 29.68%]), exposes an action on 18.99% of pre-onset prefixes, and predicts 5.94% of complete trajectories exactly. Its pooled label score is at the 96th percentile of 100 within-prefix label permutations, below the 97.5th-percentile reference (26.73%). It exceeds matched text, scalar-acoustic, and shuffled-representation probes in pooled post-onset label accuracy. Elapsed time is more onset-exact (36.25% versus 23.13%) but less accurate about action identity (9.92% versus 26.03%). These results motivate separate measurement of action identity and onset timing in partial-speech evaluations.
Read the full abstract

Audio and Speech Processing · Cross-listed from cs.SD

Enabling Immersive Audio-Visual Experience from Any Video

Zitong Lan, Mutian Tong, Jiatao Gu et al. (+1 more)

Most videos capture only a narrow field of view and provide no spatial audio, limiting the sense of immersion they can provide. Recent video generation models can expand perspective videos into panoramic ones, but do not provide the corresponding spatial soundscape. Without spatially consistent audio, these expanded visual worlds remain incomplete. This paper presents OmniDream, a training-free framework that transforms a silent monocular video into an immersive audiovisual experience, where viewers can freely look around while sounds remain spatially aligned with the visual scene. At the core of OmniDream is an object-centric audio representation that disentangles each sound source's intrinsic audio content from its scene-dependent acoustic effects, enabling independent audio generation, physics-based simulation of propagation effects, and flexible spatial audio rendering. Experiments show improved audio-visual alignment, spatial correctness, and perceptual immersiveness over baselines. Examples are available on https://huggingface.co/spaces/CuriousAlien000/spatial-audio-360-demo
Read the full abstract

All authors Zitong Lan, Mutian Tong, Jiatao Gu, Mingmin Zhao

Audio and Speech Processing

InstCharVoice: Grounding Natural-Language Instructions for Character-Level Control in Text-to-Speech

Sihang Nie, Xueru Li, Xiaofen Xing et al. (+4 more)

Instruction-based text-to-speech (ITTS) systems enable natural-language control of expressive speech generation, but often offer limited transparency and fine-grained control over individual text units. Character-level controllable TTS systems provide explicit acoustic control, yet typically rely on user-specified acoustic attributes. To bridge this gap, we propose InstCharVoice, a unified framework that grounds natural-language instructions in character-level acoustic control. We first construct grounded instruction annotations on the WordVoice-5A-zh corpus using Qwen3-Omni. With this supervision, we train an autoregressive model to identify instruction-relevant characters and predict their acoustic attributes before generating the corresponding speech tokens. Keyword prediction and grounding-aware loss weighting help the model focus on instruction-relevant characters and attributes. Experiments show improved instruction following and keyword-level acoustic control over representative ITTS systems, with competitive speech naturalness and explicit character-level controllability. Audio samples are available at https://xxh333.github.io/instcharvoice-demo/.
Read the full abstract

All authors Sihang Nie, Xueru Li, Xiaofen Xing, Deyi Tuo, Cheng-Bin Jin, Jingyuan Xing, Jinxin Ji

Audio and Speech Processing

Model-Guided Design of Low-Context Speech Probes for Cochlear Synaptopathy

Ahsan J. Cheema, David Meng, Jorge Mejia et al. (+2 more)

Cochlear neural degeneration (CND) can impair suprathreshold coding without elevating pure-tone thresholds, complicating its diagnosis when it coexists with hair cell loss. We present a unified comparison of temporal and noise-based probes for CND detection using low-context vowel-consonant-vowel (VCV) syllables to reduce linguistic and contextual cues. Using a phenomenological auditory nerve model, we simulated responses to 21 VCV tokens under time compression, reverberation, and speech-in-noise conditions across presentation levels and seven CND profiles. We computed mutual information (MI) between inner hair cell potentials and auditory nerve neurograms and quantified information loss relative to a normal-hearing baseline. Time compression and amplitude-modulated (AM) noise produced the largest modeled information losses. We then evaluated these stimuli in a consonant-identification study involving 36 listeners with normal audiograms, 12 of whom reported difficulty understanding speech in noise. Neither 40 percent time compression in quiet nor AM noise alone distinguished listeners with and without these difficulties. However, compressed speech presented in AM noise separated the two groups. This partial agreement between model predictions and behavior supports our MI-based stimulus design framework and motivates further evaluation of combined temporal and noise-based probes for CND detection.
Read the full abstract

All authors Ahsan J. Cheema, David Meng, Jorge Mejia, Sanna Hou, Sunil Puria

Audio and Speech Processing · Cross-listed from cs.SD

CoSE-E: A Benchmark for Code-switched Speech Evaluation in Enterprise Settings

Shama Gupta, Hoang H Nguyen, Chelsea Huang et al. (+2 more)

Code-switching (CS), a seamless alternation between languages within a single utterance, remains a critical challenge in automatic speech recognition (ASR). While prior works focus on conversational CS-ASR, enterprise settings demand evaluation of operational impact beyond edit-distance errors: how code-switching transcription errors propagate to downstream voice agent task failures. In this work, we propose (1) a CS-ASR synthetic benchmark and multidimensional evaluation framework tailored to enterprise domains, (2) systematic evaluation of frontier ASR systems across 5 language pairs, (3) diagnostic analysis of the additional transcription errors that code-switching introduces across language pairs and models. We release COSE-E to support enterprise-focused CSASR evaluation for multilingual voice agents in enterprise deployment.
Read the full abstract

All authors Shama Gupta, Hoang H Nguyen, Chelsea Huang, Lindsay Devon Brin, Fanny Riols

Audio and Speech Processing

Simulation-Based Inference for Plate Reverb System Identification

Dylan Sechet, Marc Evrard and Matthieu Kowalski

We address Task A of the 1st DAFx Parameter Estimation Challenge, which aims to retrieve the physical parameters of a plate model from an impulse response. To do so, we use the Simulation-Based Inference (SBI) framework, in which we train a neural network to estimate a density over plate parameters given an impulse response, using a dataset generated by the simulator. Inference for a new impulse response then requires only a forward pass through the network, without involving the simulator. For each test observation, we fine-tune a specific network: additional simulation rounds are performed by sampling parameters from the current estimated distribution, simulating the corresponding impulse responses, and fine-tuning to produce the specialized network.
Read the full abstract

Audio and Speech Processing

Perceptual Quality Loss or Loss of Perceptual Quality?

Danilo de Oliveira, Tal Peer, Maurício do V. M. da Costa et al. (+1 more)

Contemporary deep speech enhancement (SE) models are often trained with specific auxiliary terms in the loss function as a way to improve their performance in terms of perceptual metrics. Nevertheless, a higher score on a perceptual metric does not necessarily correlate with an improved listening experience. Through objective and subjective experiments, we assess the performance of SE models trained with two different types of auxiliary PESQ loss terms. The numerical evaluation on a suite of standard metrics suggests that, while models optimized for PESQ naturally obtain higher PESQ scores in the test set, for most other metrics the scores do not significantly change. In some cases, the PESQ loss even results in worse PESQ scores on mismatched data. A formal listening experiment reveals that the models without a PESQ loss were generally preferred over models that include it, across all settings. Finally, we analyze the relative importance of PESQ in the composite metrics CSIG, CBAK and COVL, and find that PESQ dominates all of them. Our study highlights the perils of over-reliance on PESQ and stresses the importance of a complete evaluation procedure for SE.
Read the full abstract

All authors Danilo de Oliveira, Tal Peer, Maurício do V. M. da Costa, Timo Gerkmann

Audio and Speech Processing · Cross-listed from cs.SD

JazzSAMBA: A Synchronous and Asynchronous Multi-take Band Audio Dataset of Jazz Standards for Live Music Models

Phillip Long, Jacob Nguyen, Jace Hosto et al. (+9 more)

Machine learning has made strong progress on music tasks, both as assistive tools and as creative partners. However, most systems train on multitrack corpora that emphasize pop and rock. Jazz, with improvisation at the core of its practice, still lacks a well-annotated corpus of clean per-stem combo recordings on standards. We introduce JazzSAMBA (Jazz Synchronous and Asynchronous Multi-take Band Audio) to fill this gap: the first originally recorded jazz-combo multitrack dataset of standards with asynchronous (overdubbed) and synchronous (live ensemble) protocols, preferred and alternate takes chosen by the musicians, and timed annotations for bars, chords, sections, and soloists. JazzSAMBA covers 76 standards by eight musicians on drums, bass, piano, trumpet, and saxophone, with per-stem audio, mixtures, and MIDI. It can support chart-conditioned accompaniment, combo source separation, and form-aware music information retrieval. We demonstrate the dataset on two tasks: a jazz combo source-separation baseline and a chart-conditioned accompaniment ablation. The dataset, code, and samples are linked from the project demo page.
Read the full abstract

All authors Phillip Long, Jacob Nguyen, Jace Hosto, Gage Hosto, Jett Takazawa, Fares Nofal, Sebastian Stade, Nithya Shikarpur, Julian McAuley, Cheng-Zhi Anna Huang, Stephen Brade, Aleksandra Teng Ma

Audio and Speech Processing

Domain-Incremental Learning for Generative Speech Enhancement

Manjunath Mulimani, Annamaria Mesaros, Minje Kim et al. (+1 more)

We propose a domain-incremental learning framework for generative speech enhancement (SE) that learns from a sequence of datasets or domains recorded under diverse acoustic conditions. Fine-tuning a pretrained model on continuously evolving domains leads to catastrophic forgetting of previously acquired knowledge, while zero-shot generalization often fails to adequately adapt to unseen domains. To address these challenges, we first develop a novel language model-based generative SE model that we then use as a pretrained backbone and incrementally adapt it to acoustically mismatched domains using lightweight domain-specific Low-Rank Adaptation. The proposed framework enables the model to acquire enhancement capabilities for new domains while preserving performance on previously learned domains. Evaluated on four heterogeneous speech datasets, our approach effectively adapts to new domains without forgetting previously learned domains.
Read the full abstract

All authors Manjunath Mulimani, Annamaria Mesaros, Minje Kim, Jesper Rindom Jensen

Audio and Speech Processing · Cross-listed from cs.SD

Unsupervised Speech Enhancement via Drifting

Diego Caviedes-Nozal, Liang Xu, Rasmus Kongsgaard Olsson et al. (+1 more)

This paper addresses unsupervised speech enhancement in the unpaired setting using drifting methods, where training relies on separate collections of degraded and clean audio without corresponding pairs. While recent drifting approaches enable unpaired training, they do so at a heavy cost: because the objective optimizes only a marginal prior over clean speech, the enhancer gradually loses the input's linguistic content and speaker identity. To fix this, we introduce input-conditioned drifting. We preserve the pull of the clean corpus while re-tethering the output to the degraded input via two mechanisms: an anchor encoder supplies the missing likelihood by pulling toward the input's features, and a key encoder conditions the prior by re-weighting retrieved frames. Neither requires labels or paired data. Using a training-free encoder selection criterion, Word Error Rate on VoiceBank-DEMAND falls to 10.1% (unprocessed: 11.7%), speaker similarity recovers from 0.490 to 0.879, and the recipe transfers in part to dereverberation on WSJ0-REVERB: content improves, rendering quality does not.
Read the full abstract

All authors Diego Caviedes-Nozal, Liang Xu, Rasmus Kongsgaard Olsson, W. Bastiaan Kleijn

Audio and Speech Processing

Measurement-Based Bitrate-Energy-Quality Analysis of Neural Audio Codec Decoders on Laptop and Phone Platforms

Seunghyeon Shin and Seokjin Lee

Neural audio codecs can achieve similar objective quality at lower bitrates than conventional codecs, but their decoder-side computational cost may offset this bitrate advantage on battery-powered client devices. This paper presents a measurement-based rate–energy–quality analysis of four neural audio codecs–EnCodec, DAC, HILCodec, and SNAC–and two conventional baselines, AAC-LC and Opus, on laptop and phone platforms. Speech and music are evaluated separately using the original-reference ViSQOL protocol. For the main comparison, operating points are matched by selecting the measured point nearest to the midpoint of the common overlap in treatment-level mean ViSQOL. A secondary analysis includes only explicitly evaluated bitrate settings. Decoder energy is measured as idle-subtracted energy per second of audio (J/s), and results are summarized as the median of three repeated runs for execution-valid runtime and device paths. Pairwise break-even transmission-energy thresholds are then derived analytically from the measured decoder energy and bitrate. At the matched-quality operating points, the evaluated neural codecs achieved similar ViSQOL scores at lower bitrates but generally required more decoder-side energy than the applicable conventional codecs. EnCodec produced the lowest neural break-even thresholds in both matched-quality cohorts on both platforms. By contrast, several DAC and SNAC comparisons on the Phone XNNPACK CPU path exceeded 200 mJ/kbit, and their full-band Phone configurations required more than 1 s to decode 1 s of audio. These results show that lower bitrate alone does not guarantee an energy benefit: deployment efficiency also depends on decoder complexity, the effective runtime and device mapping, execution validity, accelerator availability, and the transmission-energy coefficient.
Read the full abstract

Audio and Speech Processing

CharDuplex: Building Character-Consistent Full-Duplex Spoken Dialogue Models

Donghang Wu, Yisi Liu, Chen Chen et al. (+2 more)

Full-duplex speech models are moving voice interaction beyond conventional turn-taking, yet natural conversation is shaped not only by when an agent speaks, but also by how it behaves as a conversational character. We present CharDuplex, a character-driven full-duplex speech model that combines real-time spoken interaction with persona-conditioned behavior. We first adapt GLM-4-Voice to an always-on dual-stream architecture and train the model for full-duplex conversation. Then a fully automated pipeline constructs character-conditioned dialogue data from open-source character descriptions for character-conditioned supervised fine-tuning. The model is further refined with the proposed FDGym, where an LLM-simulated user dynamically interacts with the model, enabling reinforcement learning over evolving multi-turn interactions. On SpeechRole-Eval, CharDuplex achieves the highest average score among the evaluated open-source models, while remaining competitive with closed-source systems. It also demonstrates competitive general speech intelligence and strong full-duplex interaction capabilities. CharDuplex demonstrates a practical training recipe for building full-duplex voice assistants that are not only interactive, but also character-consistent.
Read the full abstract

All authors Donghang Wu, Yisi Liu, Chen Chen, Hexin Liu, Eng Siong Chng

Audio and Speech Processing

Beyond Textual Chain-of-Thought: JEPA-Conditioned Latent Reasoning for Large Audio Language Models

Donghang Wu, Haoyang Zhang, Yizhou Peng et al. (+6 more)

Explicit textual Chain-of-Thought (CoT) has improved the reasoning ability of large audio language models (LALMs). However, textual CoTs are often constructed from text captions of audio and provide limited access to the acoustic evidence, which can introduce problems like hallucination. To address this modality-gap issue, we introduce JELAR, a Joint-Embedding Predictive Architecture (JEPA)-based latent reasoning framework that conditions latent reasoning supervision on acoustic representations learned from raw waveforms. During training, a frozen WavJEPA model provides representations learned directly from raw waveforms. A non-causal expert first constructs answer-aware queries, which cross-attend to WavJEPA embeddings to produce latent reasoning targets. The LALM is trained to predict these targets before generating its response. Experimental results show that JELAR improves the Audio-Reasoner baseline by 2.70 and 9.10 absolute percentage points on MMAU-mini and MMAR, respectively, demonstrating the effectiveness of JEPA-conditioned latent reasoning as an alternative to explicit textual CoT supervision.
Read the full abstract

All authors Donghang Wu, Haoyang Zhang, Yizhou Peng, Shreyas Gopal, Yi-Wen Chao, Chen Chen, Hexin Liu, William Tjhi, Eng-Siong Chng

Audio and Speech Processing · Cross-listed from cs.CV

Joint and Cross-Modal Video-Audio Generation and Editing: A Unified Formulation and Design Taxonomy

Abhinav Sharma, Sai Karthik Navuluru, Wang Wei et al. (+21 more)

Video and audio are perceived together, yet most generative models treat them in isolation. We examine methods that model the two modalities jointly, generate one from the other, or edit them in a coupled manner, organized around a single question: how is the output kept coherent across modalities in time and semantics? A unified formulation casts joint generation, cross-modal generation, and joint editing as three problems defined on a single distribution over audio-visual pairs, and a taxonomy compares methods along five design axes. To our knowledge, this is the first overview to systematically taxonomize joint audio-visual editing, which we map as nine edit categories spanning 28 edit types. We describe methods, datasets, and metrics for each setting and close with the open problems we view as most consequential.
Read the full abstract

All authors Abhinav Sharma, Sai Karthik Navuluru, Wang Wei, Daksh Dangi, Xiangbo Gao, Li Li, Bo Ni, Vardhan Dongre, Junda Wu, Xiyang Hu, Jiuxiang Gu, Seunghyun Yoon, Tong Yu, Chien Van Nguyen, Mohamed Elmoghany, Nedim Lipka, Hoda Eldardiry, Hongjie Chen, Tyler Derr, Thien Huu Nguyen, Zhengzhong Tu, Nesreen K. Ahmed, Franck Dernoncourt, Ryan A. Rossi

Audio and Speech Processing · Cross-listed from cs.SD

SAIL: Spatial Audio Intelligence with Large Language Models via Disentangled Acoustic-Spatial Encoding and Dual-Stream Q-Former

Zhengding Luo, Jinyang Wu, Haozhe Ma et al. (+3 more)

Spatial audio large language models (LLMs) enable embodied agents, wearable assistants, and immersive systems to recognize sound events, localize sources, and reason about their spatial relationships. However, existing spatial audio LLMs often rely on early fusion of acoustic and spatial features and source-agnostic token representations. These designs make it difficult to preserve the correspondence between individual sound events and their spatial attributes, particularly in multi-source scenes. To address this limitation, we propose SAIL, a Spatial Audio Intelligence framework with LLMs that preserves acoustic-spatial structure and source-level correspondence from audio encoding to LLM alignment. SAIL introduces a Disentangled Spatial Audio Transformer that represents Mel-spectrogram and interaural phase difference features as separate acoustic and spatial streams. Source-discriminative task queries further learn event, direction, and distance information for each source. A Dual-Stream Q-Former then aligns the two streams with the LLM using acoustic and spatial queries organized by source slots. Compared with the early-fusion baseline, SAIL achieves consistent improvements in dual-source sound event detection, direction and distance estimation, and spatial reasoning. These results demonstrate the importance of structured, source-discriminative audio representations for multi-source spatial understanding and reasoning.
Read the full abstract

All authors Zhengding Luo, Jinyang Wu, Haozhe Ma, Yanghao Zhou, Woon-Seng Gan, Wenwu Wang

Audio and Speech Processing

Audio Tokens as a Budgeted Resource: Marginal-Utility Allocation for Scalable Audio Representations

Mingyu Zhao, Jinchao Zhang and Zhiyong Wu

Discrete audio tokens are widely used as a representation interface, yet fixed-depth RVQ tokenizers allocate equal capacity to every frame despite varying refinement value. We introduce UniAdapt, which learns marginal utility of RVQ refinements on a frozen codec and allocates them under exact serialized-bit budgets. A rate-independent causal controller predicts acoustic utility, while an optional semantic head supports speech-only utterance-level allocation; measured acoustic and semantic marginal gains on speech have a correlation of 0.42. For causal allocation, a primal-dual allocator selects prefix-valid depths, while an exact guard constrains each sequence prefix to its matched fixed-depth serialized budget. Under utterance-level allocation, UniAdapt reduces Log-STFT distortion by 1.07-4.39 percent across speech, music, and environmental audio without larger budgets. Causally, it improves three of four speech rates with zero violations across 800 utterance-rate evaluations and runs faster than real time. A 20-listener utterance-level MUSHRA study shows a significant 3.52-point speech improvement, with no significant differences on music or environmental audio. These results support separating utility prediction from budget enforcement for scalable, budget-conditioned audio representations.
Read the full abstract

Audio and Speech Processing · Cross-listed from cs.CL

SALMONN-duo: Adaptive Dual-System Coordination for Full-Duplex Voice Agents

Wenyi Yu, Siyin Wang, Terumi Chiba et al. (+5 more)

Full-duplex speech large language models (LLMs) enable low-latency, natural voice interaction. However, real-world agents must also use tools and perform deliberative reasoning-operations whose variable latency and computational cost conflict with the stringent timing requirements of real-time conversation. To reconcile these demands, we propose SALMONN-duo, an adaptive dual-system voice agent inspired by dual-process theories of cognition. SALMONN-duo separates real-time interaction from deliberative computation by pairing an always-on, fast-thinking full-duplex speech LLM (system 1) with a powerful asynchronous slow-thinking LLM agent (system 2). Beyond handling real-time interaction, system 1 learns when to answer directly and when to delegate, remaining responsive during backend execution and seamlessly integrating returned information into the ongoing dialogue without exposing tool traces or losing conversational context. Evaluations on single-turn spoken question answering (QA) and multi-turn conversations demonstrate that adaptive delegation substantially improves accuracy on knowledge-intensive and multi-hop reasoning questions, while knowledge-boundary-aware training avoids unnecessary system 2 invocations. On a customized version of $\tau$-Voice, SALMONN-duo further demonstrates its ability to complete environment-grounded, policy-constrained tasks through multi-turn interactions in realistic business scenarios. Finally, cost-aware reinforcement learning further enhances the trade-off between task performance and backend usage across the QA and conversation tasks, while improving task success and response safety on $\tau$-Voice with an acceptable increase in the delegation rate.
Read the full abstract

All authors Wenyi Yu, Siyin Wang, Terumi Chiba, Xianzhao Chen, Xiaohai Tian, Jun Zhang, Lu Lu, Chao Zhang

Audio and Speech Processing

Explainable and Generalisable LLM-based Cognitive Decline Detection with Spontaneous Speech

Ziyun Cui, Wen Wu, Chuan Shi et al. (+10 more)

Alzheimer's disease (AD) and mild cognitive impairment (MCI), which may precede AD, manifest early through subtle linguistic and acoustic alterations. Traditional diagnostics, however, are often resource-intensive and lack scalability for mass screening. To address these challenges, we introduce a novel bilingual speech large language model framework for automated, explainable cognitive screening. Unlike conventional pipelines that rely on error-prone automatic speech recognition, our system directly processes raw speech to learn joint acoustic-semantic representations, preserving critical prosodic cues often lost in transcription. Utilising our newly collected PUTH-AD dataset alongside multiple open-source corpora, we implemented a multi-task learning objective that simultaneously performs cognitive status classification and generates clinician-understandable natural language explanations. Our system achieved the highest average accuracy and AUROC across six dataset/task conditions, comparing three representative baselines. The system demonstrated cross-task transfer to held-out PUTH-AD task subsets, maintaining classification accuracy on an entirely unseen cognitive task without task-specific fine-tuning. Furthermore, clinician evaluation confirms that the generated explanations are both clinically relevant and largely consistent with the underlying speech evidence, supporting their potential utility in clinical interpretation. This study provides a scalable, objective, and explainable framework for speech-based cognitive screening, combining cognitive status classification with natural language explanations that clinicians can assess and verify, bridging the gap between advanced AI and clinical utility.
Read the full abstract

All authors Ziyun Cui, Wen Wu, Chuan Shi, Shuguang Yang, Xueying Gui, Yan Zheng, Qiong Yang, Haiyan Zhao, Wei-Qiang Zhang, Ji Wu, Yelei Li, Nan Li, Chao Zhang

Audio and Speech Processing

GAMF: Learned and Analytical Array Transfer Function Matching for Array-Generic Direction-of-Arrival Estimation

Zhiheng Jin, Shichao Hu, Chunyang Xu et al. (+1 more)

Microphone positional encoding supports cross-array direction-of-arrival (DOA) estimation, but coordinates alone cannot fully describe device shadowing or microphone directivity. We propose a Generalizable ATF Matching Framework (GAMF) for DOA estimation across array geometries and microphone counts, using array transfer functions (ATFs) as acoustic descriptors. The learned branch incorporates ATF embeddings into geometry-conditioned neural estimation to match acoustic observations with candidate directions. The analytical branch performs normalized ATF matching adapted from generalized steered response power. A hybrid configuration combines their scores through adaptive gating. Simulations across array configurations show that both learned and hybrid configurations outperform a representative positional-encoding-based neural baseline, remain competitive with analytical ATF matching in clean, low-reverberation scenes, and substantially improve upon it under stronger noise or reverberation. On eight-microphone LOCATA Task 1 recordings, the hybrid configuration outperforms the evaluated state-of-the-art baselines for three-dimensional DOA estimation, achieving mean errors of $3.76^\circ$ for three-dimensional DOA and $2.87^\circ$ for azimuth.
Read the full abstract

All authors Zhiheng Jin, Shichao Hu, Chunyang Xu, Mengyao Zhu

Audio and Speech Processing

SPEAR-Gen: Generation-Aware Pre-training for Unified Speech Representations

Xiaoyu Yang, Arthur Hinsvark, Antonios Alexos et al. (+3 more)

Speech understanding and generation place different demands on speech representations, and existing models are typically optimised towards one capability or the other. To reduce this gap, we introduce SPEAR-Gen, a speech representation model that learns a single representation for both capabilities. Task-aligned feature aggregation consolidates complementary linguistic and paralinguistic information across a frozen encoder into discrete targets for masked prediction, while a coarse-to-fine objective combines log-Mel reconstruction with residual flow matching to preserve spectral structure and fine-grained acoustic variation. Experiments on SUPERB and speech resynthesis show that SPEAR-Gen maintains strong understanding performance while substantially improving resynthesis quality and speaker preservation. These results demonstrate that a single speech representation can effectively support both understanding and generation.
Read the full abstract

All authors Xiaoyu Yang, Arthur Hinsvark, Antonios Alexos, Osama Hanna, Philip C. Woodland, Yiting Lu

Sunday 27 September

Audio and Speech Processing

Is Word Error Rate Enough? Rethinking Privacy Evaluation in Speech with Entity-Aware Metrics

Anjana Rajasekhar, Jule Pohlhausen, Nayana Jacob Alappattu et al. (+1 more)

As the use of smart devices continues to increase, their potential to capture sensitive speech content raises growing privacy concerns. It is therefore critical to develop techniques that prevent information leakage while preserving the utility of the audio, and evaluation metrics that accurately quantify the level of privacy without overestimating it. In this work, we evaluate the effectiveness of two obfuscation techniques in protecting speech content, with particular emphasis on named entities, by adapting entity-aware privacy metrics from the Natural Language Processing field to the speech privacy domain. Further, we investigate several attack scenarios and show that fine-tuning on entity-rich data improves attack performance for some entity categories but not others. Finally, we provide guidance on metric selection based on whether the obfuscation method preserves temporal alignment.
Read the full abstract

All authors Anjana Rajasekhar, Jule Pohlhausen, Nayana Jacob Alappattu, Anna Leschanowsky

Audio and Speech Processing

Rethinking Automated Voice Similarity by Shifting from EER to Embedding Geometry

Szu-Chi Chen, Jia-Kai Dong, Yi-Cheng Lin et al. (+2 more)

Speaker verification (SV) models are commonly assumed to better capture nuances among speaker characteristics as verification accuracy improves, leading to their widespread use as automated proxies for human voice similarity in speech generation tasks. However, by establishing a human perceptual alignment metric and conducting systematic analysis, we demonstrate that perceptual alignment is governed far more by how a model is trained (its learning objective) than by how well it performs (EER). Notably, standard margin-based classification losses (e.g., AAM-Softmax) yield substantially lower perceptual alignment than prototypical metric losses, while EER itself fails to track human judgment, directly challenging the community's implicit assumption. We trace this divergence to embedding geometry, where a model's effective dimensionality ($d_{\mathrm{eff}}$) tracks perceptual alignment with a $-0.95$ rank correlation, revealing that the dimensional spread favored by classification losses fundamentally clashes with the low-dimensional nature of human voice perception. Imposing a dimensionality bottleneck compresses $d_{\mathrm{eff}}$ and raises perceptual alignment ($\rho_{\mathrm{align}}$) from 0.08 to 0.74, establishing a principled geometric criterion for evaluating voice similarity.
Read the full abstract

All authors Szu-Chi Chen, Jia-Kai Dong, Yi-Cheng Lin, Sung-Feng Huang, Hung-yi Lee

Audio and Speech Processing · Cross-listed from cs.CL

In-Context Adaptation of Encoder-Decoder Models in Speech Recognition

Yen Meng, Sharon Goldwater and Hao Tang

In-context learning offers an appealing approach to adapt automatic speech recognition (ASR) models to new speakers, accents, and domains by providing speech-text pairs as demonstrations at inference time. Recent work shows that some LLM-based speech models are capable of ASR in-context adaptation, when providing interleaved speech-text demonstrations. In this work, we ask whether in-context adaptation is an inherent ability for all encoder-decoder models. We study two forms of demonstration, collated and interleaved demonstration, across six encoder-decoder models, spanning conventional cross-attention-based and LLM-based architectures. We find that all tested models are able to perform in-context adaptation out of the box, achieving up to 30% relative improvement in the oracle experiments and up to 23% using first-pass hypotheses. Through controlled experiments on three English datasets, we show that lexical and speaker information both contribute to successful adaptation. While interleaved demonstration is effective in certain cases, collated demonstration brings consistent adaptation across the board. Our results suggest that in-context adaptation for ASR is not unique to specific architectures, training, or demonstration approaches.
Read the full abstract

Audio and Speech Processing · Cross-listed from eess.SP

Unified Target-Speaker ASR with Text and Enrollment Speech Cues

Yuxiang Mei, Yuchen Yan, Dongxing Xu et al. (+2 more)

Target-speaker automatic speech recognition (TS-ASR) aims to recognize a designated speaker while suppressing interfering speech in multi-talker environments. Conventional TS-ASR typically relies on an enrollment utterance, whereas text-guided methods use known lexical content, such as a wake word, to identify the target speaker from the observed mixture. These two cues provide complementary information but are usually studied separately. We propose a Unified Dual-Cue TS-ASR framework that supports text cues, enrollment speech, or both within a single model. Text cues interact with the mixture representation to extract target-speaker information conditioned on known lexical content, while an independent enrollment utterance provides complementary speaker information. Cross-attention cue-conditioning modules are integrated into shared Conformer blocks, and negative-cue sampling provides cue-validity supervision during dual-cue training. Experiments on 30,000 two-speaker mixtures across five recording/domain conditions and four oracle text-cue lengths show that, with five-character text cues, the concatenated dual-cue method achieves 8.80% CER, compared with 17.32% for text-only and 29.06% for enrollment-only inference. It also outperforms parallel dual-cue fusion (9.49% CER) and yields lower dual-cue CER across all five evaluation subsets. These results demonstrate the benefit of jointly exploiting complementary lexical and speaker information for target-speaker ASR.
Read the full abstract

All authors Yuxiang Mei, Yuchen Yan, Dongxing Xu, Jiaen Liang, Yanhua Long

Audio and Speech Processing · Cross-listed from cs.SD

Controlling Speaking Rate in Autoregressive TTS via Activation Steering

Francesco Verdini, Antonis Asonitis, Aref Farhadipour et al. (+4 more)

Autoregressive text-to-speech (TTS) systems synthesize natural speech but, once trained, offer little control over speaking rate. We show that speaking rate can be steered at inference time, without retraining, by clamping a single decoder block's activation along a discovered speed axis. A decoder-block analysis recovers the rate axis, a neutral operating point, and a per-step intensity scale; at inference, the activation's projection onto this axis is set to a fixed scalar. Learning this direction from synthetically time-stretched and time-compressed speech yields rate control that largely preserves speaker identity, generalizes across model architectures, and maintains high naturalness in objective and human evaluations. Unlike standard additive steering, which breaks at the slow extreme, clamping remains stable on all three systems tested; at moderate targets, the better rule depends on the model. Finally, we show that rate information is decodable across layers but causally steerable only within a mid-depth window, and demonstrate the effectiveness of our approach on the public Seed-TTS-Eval benchmark.
Read the full abstract

All authors Francesco Verdini, Antonis Asonitis, Aref Farhadipour, Marzieh Razavi, Pierre-Edouard Honnet, Vijeta Avijeet, Juan Pablo Zuluaga Gomez

Audio and Speech Processing · Cross-listed from cs.SD

Tokens Change, Structure Endures: Spectral Watermarking for Generated Speech

Kanghwi Lee, Kyeongseok Jeong and Jeongmin Liu

Watermarking is a promising tool for establishing the provenance of AI-generated speech. While many neural audio watermarking methods rely on a separately trained watermark generator, token-level watermarking is a training-free alternative that operates directly during generation. Its main weakness is retokenization: decoding generated speech to a waveform and encoding it again can change token identities and erode the watermark. To make the watermark robust to these changes, we propose Redwing, REtokenization-Durable Watermarking IN Generation. It builds a graph from the token substitutions observed under retokenization, whose Laplacian yields a basis that assigns similar values to tokens likely to substitute for one another. Over this basis, embedding and detection functions are jointly optimized to preserve watermark signal through retokenization while limiting embedding distortion and detector variability on unwatermarked speech. On the Moshi full-duplex system, after eight consecutive passes of Mimi resynthesis, Redwing achieves 80.7% TPR at a calibrated 1% FPR, compared with 8.3% for KGW and at most 7.3% for WMAR. It also has the highest TPR after eight passes through three other neural codecs (77.5-93.0%), and the gains generalize to TTS models at a speech-quality cost close to that of KGW. These results show that retokenization is not merely a source of noise: its transition structure can be exploited as a design principle for robust token-level watermarking.
Read the full abstract

Audio and Speech Processing

YuE2: Unifying Symbolic and Audio Music Generation at Frontier Quality

Ruibin Yuan, Jiahao Pan, Junyan Jiang et al. (+32 more)

Symbolic models make melody, harmony, rhythm, and form explicit but typically stop before a finished recording; audio models produce complete songs while leaving composition implicit. We introduce YuE2, which unifies symbolic and audio music generation at frontier quality through symbolic planning. A single AR-NAR Mixture-of-Transformers (MoT) first writes a readable score specifying melody and harmony, expands it into semantic music tokens, and realizes it as full-song audio. In comparisons using the same checkpoint, experts prefer symbolic planning for overall quality and musicality, with 49.3% of overall preferences versus 34.6% without planning. Experts also favor the unified model over a separate language model and diffusion Transformer. On WildSongBench, YuE2 scores 6.73 on SongBench Global Avg, exceeding all evaluated public baselines. Selecting from eight candidates (best-of-8), YuE2 reaches 6.96, the highest observed mean among all evaluated systems. Expert listening further establishes its competitiveness with proprietary song generators, favoring best-of-8 over Suno v4.5 and yielding nearly balanced preferences against Suno v5. To learn this generation process from recordings without aligned scores, we introduce MERT2 and SheetSage2 to supply semantic and symbolic supervision. MERT2 sets a new state of the art in music representation learning, surpassing previous best results on 14 of 15 MARBLE metrics; SheetSage2 leads 12 of 15 benchmark-metric pairs in our lead-sheet transcription comparison. The same checkpoint follows score edits while largely preserving unedited musical content and generates zero-shot covers without cover-specific training. Its readable score also enables agentic music editing, with external language models translating user feedback into revisions of the composition.
Read the full abstract

All authors Ruibin Yuan, Jiahao Pan, Junyan Jiang, Zhiyue Wu, Ziya Zhou, Jiankai Sun, Yizhi Li, Ge Zhang, Yicheng Gu, Zeyue Tian, Junyu Dai, Hanfeng Lin, Kai Li, Shangda Wu, Xuanjie Liu, Jiaming Wang, Zihan Liu, Yue Wang, Yinghao Ma, Hanzhi Yin, Kangrui Chen, Xinyue Zhang, Ziyang Ma, Mengqi Liao, Hejia Zhao, Guowei Huang, Chao Yan, Lei Ke, Jianwei Yu, Bei Liu, Joe Guo, Liumeng Xue, Gus Xia, Wei Xue, Yike Guo

Audio and Speech Processing · Cross-listed from cs.SD

Transformer-based Neural Beamforming for Real-Time Speech Enhancement on Smart Low-Power Hearable Devices

Luca Bompani, Marco Fariselli, Giovanni Oltrecolli et al. (+1 more)

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.
Read the full abstract

All authors Luca Bompani, Marco Fariselli, Giovanni Oltrecolli, Francesco Conti