Peer Revenue Fully funded

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

Monday 17 August

Audio and Speech Processing

Contrastive Learning with Variational Regularization for Multi-Session EEG-to-Speech Decoding

Tomoaki Mizuno and Toru Nakashika

Reconstructing heard speech from non-invasive electroencephalography (EEG) is challenging due to a low signal-to-noise ratio (SNR) and inter-session variability. While trial averaging improves the SNR, it is difficult to apply to continuous speech. We instead use repeated EEG responses to the same stimulus across different sessions as positive pairs for contrastive learning, and introduce variational regularization that, combined with this contrastive objective, keeps the encoder representation space broad. Experiments on a Japanese EEG dataset show that combining the session-invariant strategy with variational regularization improves the character error rate (CER) while maintaining mel-spectrogram reconstruction fidelity. Session probing confirms that the encoder representations achieve session-invariance.
Read the full abstract

Audio and Speech Processing

A Novel Binaural Cue Preservation Loss for DNN-Based Binaural Speech Enhancement

Jayteerth Amble, Thomas Haubner, Hendrik Schröter et al. (+2 more)

Binaural speech enhancement for hearing aids aims to reduce noise while preserving the interaural cues needed for spatial localization. Although deep neural network-based methods achieve strong noise reduction, they often distort the rela- tionship between the left and right signals. In this paper, we propose two novel binaural cue preservation losses. First, a binaural reconstruction error loss that directly penalizes masking-induced distortion in the relationship between the left and right spectra, providing a more direct measure of the binaural consistency than conventional separate interaural level differences (ILD) and interaural phase differences (IPD) errors as in prior work. Second, a binaural cue loss that jointly models ILD and IPD to better preserve the binaural structure. Experimental results show that both proposed losses maintain strong noise reduction performance and reduce masking- induced distortion compared to the state-of-the-art baseline cue loss, while the second proposed joint binaural cue loss also outperforms the baseline in ILD preservation.
Read the full abstract

All authors Jayteerth Amble, Thomas Haubner, Hendrik Schröter, Christoph Hoog Antink, Henning Puder

Audio and Speech Processing

Geometry-adaptive Ambisonic encoding for sparse microphone arrays of variable topology using physics-informed diffusion

Xiang Zhou, Zhengqiao Zhao, Zhengding Luo et al. (+1 more)

Ambisonics delivers compact scene based spatial audio representation, yet higher order Ambisonic encoding poses difficulties for wearables and embedded hardware. Their microphone arrays are often sparse, irregular, and constrained by device specific boundary conditions. These factors make the spherical-harmonic (SH) domain encoding ill conditioned: inverse filtering amplifies noise, while deterministic neural encoders may overfit to array-specific responses or smooth ambiguous higher-order components. This paper presents DiffM2A, a geometry-adaptive conditional diffusion framework for robust Ambisonic encoding from sparse MAs with variable topologies. Its Geometry-Adaptive Spherical Harmonic Projection (GASHP) front-end constructs boundary-aware SH steering functions and applies an energy-normalized modal projection, mapping array-dependent observations to a common modal representation without explicit pseudo-inverse computation. A dual-branch Elucidated Diffusion Model then estimates complex Ambisonic coefficients, conditioned on both the raw microphone spectra and GASHP features. Sound intensity and rotational equivariance losses further enhance inter-channel phase consistency and structured behavior across SH subspaces. Evaluations on both first- and second-order Ambisonic encoding tasks, using simulated room-acoustics and real-world LOCATA recordings, demonstrate that DiffM2A outperforms conventional and neural baseline methods on signal fidelity, spectral accuracy, spatial coherence, and binaural cue preservation. Additional experiments show that these gains are largely retained across unseen five-microphone layouts and under mismatched open-array and rigid-sphere boundary models.
Read the full abstract

All authors Xiang Zhou, Zhengqiao Zhao, Zhengding Luo, Wen Zhang

Audio and Speech Processing

NormToken: Speaker- and Duration-Normalized Semantic Speech Tokens via Iterative S2U-T2U Refinement

Hanlin Zhang, Daxin Tan, Dehua Tao et al. (+3 more)

Semantic speech tokens should preserve linguistic content while suppressing utterance-specific acoustic and duration variation. However, existing speech-to-unit (S2U) tokenizers often retain speaker-related acoustic characteristics and duration information. To address this issue, we propose NormToken, an iterative semantic token purification framework that alternates S2U and text-to-unit (T2U) training. In each iteration, the T2U model produces text-derived tokens to supervise a newly initialized S2U tokenizer. The resulting S2U tokens are then used as targets for the next T2U iteration. This cycle drives the two models toward a shared, text-predictable token space. Experiments on Mandarin and English demonstrate improved S2U–T2U agreement and parallel-utterance token consistency. De-tokenizers trained on the initial and refined tokens further show that refined tokens maintain comparable WER and CER while improving speaker similarity in both voice cloning and text-to-speech synthesis. In voice cloning, refined tokens produce speaking rates closer to the acoustic reference, suggesting reduced dependence on source duration information. Audio samples are available at https://hanlin1004.github.io/normtoken_demopage.
Read the full abstract

All authors Hanlin Zhang, Daxin Tan, Dehua Tao, Chengxi Deng, Xiao Chen, Linqi Song

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

INSPIRE: A Benchmark for Instruction-Aware Speech Retrieval

Chen-An Li and Hung-yi Lee

Existing speech retrieval systems rely on fixed similarity matching and cannot adapt to diverse user intents. We introduce INSPIRE, the first benchmark for instruction-aware speech retrieval, in which natural-language instructions dynamically specify relevance criteria, including semantic content, speaker identity, speaking style, environmental sounds, and their combinations. We evaluate four retrieval paradigms: large audio-language models, cascaded pipelines, self-supervised speech models, and contrastive audio-language models. Our results reveal that no current method robustly handles all retrieval intents. Text-based approaches perform relatively better at semantic retrieval but struggle with paralinguistic attributes, while speech-based models are moderately better at capturing acoustic properties but falter at following instructions. These findings highlight the need for unified architectures capable of instruction-aware speech retrieval.
Read the full abstract

Audio and Speech Processing

Navigating Speech Enhancement for Real-Time MRI: A Systematic Assessment of Signal Quality, Source Preservation, and Downstream Tasks

Huang-Cheng Chou, Sean Foley, Haley Hsu et al. (+10 more)

Audio recorded during real-time magnetic resonance imaging (rtMRI) is heavily contaminated by scanner noise, but it remains unclear whether general-purpose speech enhancement improves the signal for speech research and downstream processing. Three off-the-shelf systems—Denoiser, PASE, and RE-USE—are evaluated across five rtMRI corpora using naturally recorded inputs, a clean-input probe, and an archived paired additive-noise probe. The multi-task evaluation spans learned quality predictors, speaker and phone representations, reference-based intelligibility and quality measures, acoustic–phonetic probes, automatic speech recognition (ASR), and paralinguistic tasks. The central result is that enhancement effects are endpoint dependent: higher predicted-quality scores do not reliably imply better ASR performance or greater source fidelity. Across 15 corpus–recognizer comparisons using corpus-provided processed inputs, RE-USE yielded lower word-error-rate point estimates in 11, whereas Denoiser yielded higher estimates in 13. In the paired additive-noise probe, PASE and RE-USE improved recognized-phone agreement, intelligibility, and perceptual-quality point estimates. Denoiser improved recognized-phone agreement and short-time objective intelligibility (STOI) but reduced speaker-embedding similarity. No system was uniformly best across corpora, recognizers, and endpoints. Enhanced rtMRI audio should therefore be treated as a task-specific transformed derivative rather than a universally improved replacement for the original or DSP-processed waveform.
Read the full abstract

All authors Huang-Cheng Chou, Sean Foley, Haley Hsu, Kevin Huang, Szu-Jui Chen, Rong Chao, Louis Goldstein, Khalil Iskarous, Dani Byrd, Yu Tsao, Sudarsana Reddy Kadiri, John H. L. Hansen, Shrikanth Narayanan

Audio and Speech Processing

Feedforward Active Speech Suppression Based on Time Series Prediction of Speech Signals Using Neural Networks

Manami Nishikata and Shoichi Koyama

A feedforward active noise control (ANC) method based on time-series prediction for speech signals is proposed. Although current ANC techniques are highly effective against stationary noise, suppressing highly non-stationary speech signals remains a challenging task. We propose an adaptive filtering algorithm for active speech suppression based on neural-network-based time-series prediction of future signals. The update value for the linear control filter is calculated based on the predicted signal, as well as the current and past signals. Numerical experiments indicated that the noise reduction can be improved in both cases: when using the true predicted signal and when using a signal predicted by neural networks.
Read the full abstract

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

Agentic-DuplexGen: Decoupling Content, Timing, and Acoustics for Synthetic Dialogue Speech

Pengcheng Wang, Sheng Li, Jiyi Li et al. (+1 more)

Synthetic conversational speech has become an important resource for developing and evaluating conversational speech systems. However, existing dialogue synthesis pipelines typically generate dialogue content first and then insert interruptions, overlap, and backchannels using handcrafted markers or timing rules, making conversational timing prescribed rather than interaction-driven. We present Agentic-DuplexGen, a dialogue synthesis framework that explicitly decouples content, timing, and acoustics. An LLM first generates the dialogue script, and then two full-duplex conversational models perform the script while listening to each other in real time. This allows conversational timing to emerge naturally while preserving the scripted content. Finally, a high-fidelity text-to-speech model re-renders the interaction without altering its timing. As a demonstration of the proposed framework, we construct a patient–clinician conversational speech corpus with construction-time annotations, including word timestamps, speaker activity, overlap regions, and interaction events. Experimental results show that the proposed framework produces conversational dynamics closer to real dialogue than conventional stitching-based synthesis.
Read the full abstract

All authors Pengcheng Wang, Sheng Li, Jiyi Li, Takahiro Shinozaki

Audio and Speech Processing

Cached LLM Probability Retrieval for Speech Recognition

Sheng Li, Takahiro Shinozaki and Tatsuya Kawahara

Large language models (LLMs) enhance automatic speech recognition (ASR) by providing linguistic priors; however, their direct rescoring is costly because it requires evaluating every N-best hypothesis. This paper introduces "cached LLM probability retrieval," which involves querying a local teacher LLM offline to obtain next-token probabilities for ASR-relevant context-target pairs. These probabilities are then utilized during recognition via cache lookups, backoff strategies, and optional scoring for significant misses. The method is training-free and can integrate with existing recognizers without requiring modifications to acoustic models. Evaluations across various ASR models reveal that cached retrieval outperforms 1-pass ASR in 28 of 39 settings and achieves lower non-oracle errors. Context length analysis indicates that benefits peak at a context length of 8, suggesting that cached probability retrieval is an effective and lightweight ASR adaptation method, in contrast to the heavy training required for Generative Error Correction (GER) or knowledge distillation (KD).
Read the full abstract

Sunday 16 August

Audio and Speech Processing

Iterative Self-Learning for Expressive Text-to-Speech Synthesis

Nicholas Sanders, Gustav Eje Henter, Simon King et al. (+1 more)

Expressive text-to-speech (TTS) systems that use explicit conditioning labels provide direct and interpretable control over expressive attributes, in contrast to reference-based or prompting-based approaches, but require labeled data. Obtaining these labels at scale is costly and time-consuming, yet no prior semi-supervised framework addresses this specific bottleneck. Existing semi-supervised TTS methods instead target scarcity of paired speech-text data or transcriptions. To address the scarcity of expressive labels, we propose an Iterative Self-Learning (ISL) framework for expressive TTS, built on Invert-Classify, a classifier-free method that recovers discrete expressive labels by inverting a frozen generative model. The framework iteratively pseudo-labels unlabeled speech using the current model, retrains on the combined labeled and pseudo-labeled data, and repeats, progressively refining label quality and synthesis. We validate on two expressive tasks, word-level prominence and utterance-level emotion, across multiple low-resource data splits. We find that iterative refinement can improve pseudo-label accuracy over single-pass baselines. Furthermore, we observe that these improvements in pseudo-labeling of expressivity translate to gains in expressive label adherence and synthesis quality, confirmed by objective metrics and human listening tests. In the most data-scarce conditions, ISL-trained models outperform single-pass pseudo-labeling and further approach fully supervised performance, demonstrating that gradient-based ISL is an effective solution to expressive label scarcity in low-resource TTS.
Read the full abstract

All authors Nicholas Sanders, Gustav Eje Henter, Simon King, Korin Richmond

Audio and Speech Processing

CineDub: Scaling End-to-End Video Dubbing to Multi-Speaker Dialogues with Coherent Sound Effects

Yusheng Dai, Kangdi Wang, Baolong Gao et al. (+4 more)

Automatic video dubbing in the wild remains fundamentally limited by two competing constraints: hierarchical methods depend on brittle, multi-stage preprocessing pipelines that severely restrict data scalability and practical deployment, while holistic approaches operating on uncropped video suffer from weak temporal alignment and speaker-utterance ambiguity in multi-speaker settings. To overcome these limitations, we propose CineDub, a unified diffusion-based model that achieves precise multi-speaker dialogue dubbing directly from uncropped videos, without face cropping or speaker diarization. Central to our approach is the Implicitly-Coupled Holistic Conditioning (ICHC) paradigm, where holistic visual representations and a semantic-bundled transcription format are encoded independently, yet implicitly coupled through cross-modal training to resolve speaker ambiguity and enable precise multi-speaker multi-turn dialogue dubbing. Building on the unified temporal cues captured by holistic visual features, we further extend CineDub to joint speech and audio generation. We introduce an Ambient-to-Linguistic Curriculum Learning (ALC) to mitigate sub-task degradation, and a decoupled textual branch control mechanism to resolve cross-prompt interference during simultaneous generation. We also release two in-the-wild benchmarks, CineDub-Multi for multi-speaker dialogue dubbing and CineDub-SA for video-to-speech-and-audio (V2SA) generation, to enable evaluation under realistic conditions. Experiments show that CineDub achieves state-of-the-art results on established single-speaker dubbing and video-to-audio benchmarks while excelling in multi-speaker dialogue dubbing and acoustically coherent joint generation.
Read the full abstract

All authors Yusheng Dai, Kangdi Wang, Baolong Gao, Yuxuan Jiang, Weiqiang Wang, Qiuhong Ke, Jianfei Cai

Saturday 15 August

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

Zipf's Law of Abbreviation in a Logographic Script: Coding-Theoretic Bounds on Chinese Character Stroke Counts

Mustafa Ergen

Zipf's law of abbreviation – the tendency of frequent forms to be short – is one of the best-supported regularities in language, and recent work has moved from demonstrating it to measuring how far lexicons are compressed relative to principled baselines. That programme has so far addressed word lengths in alphabetic and syllabic scripts. We transfer it to a logographic script, taking the stroke as the unit of articulatory cost and the Chinese character as the coded form. Combining a stroke-order database covering all 20,902 characters of the CJK basic block with two independent frequency corpora (258.9M and 193.3M tokens), we find that the mean character type costs 12.71 strokes but the mean character token in running text only 7.22. Using the dually normalised optimality score of Petrini et al. (2026), the simplified inventory reaches Omega = 0.668, with the replication corpus at 0.609 – inside and just below the 62-67% band those authors report for word lengths across 20 languages and 8 scripts, suggesting a compression ceiling largely independent of script type and cost unit. A logographic script also makes absolute coding bounds computable, since strokes come from a closed five-element taxonomy: the exact 5-ary Huffman optimum is 4.34 strokes and the entropy bound 4.28, so the observed system is 1.66x above optimal coding. This gap is not slack but structure. The Kraft sum of 5^(-l_i) is 2.05 on the frequency list and 5.03 on the full inventory, so stroke strings are provably not uniquely decodable in one dimension; characters are disambiguated by the two-dimensional arrangement of strokes, not their sequence, and the forgone compression buys componential transparency. Finally, treating the mid-twentieth-century simplification reform as a controlled compression event, we find it raised optimality from 0.555 to 0.668, with savings concentrated in the 1,000 commonest characters.
Read the full abstract

Friday 14 August

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

RDSE: Real-Time Directional Speech Enhancement Framework for Dual-Mic Commodity Smartphones

Chinmay Deo, Danish Mahajan, Soudeep Ghoshal et al. (+1 more)

Audio-visual zooming on mobile devices is fundamentally constrained by limited microphone count, compute resource, unknown array geometry, and weak assumptions on real-world acoustic conditions. We present RDSE: Real-Time Directional Speech Enhancement Framework, a hybrid pipeline that combines traditional signal processing and modern machine learning methods for dual-microphone low-end smartphones. Unlike tightly coupled AV systems that continuously use visual feeds or embeddings, RDSE decouples visual interaction from feature inference, using camera input solely to parameterize the steering vector for spatial filtering. The proposed system uses four modules: dereverberation network, a beamformer with a temporal convolutional network, and a MetricGAN based denoiser. Extensive evaluation on benchmark data and real-world recordings demonstrates effective zooming without precise geometry calibration. Comparative experiments show competitive performance to existing methods. Proposed system is rigorously tested for statistical validity, ablation study, latency profiling, and practicality through real-time live-streaming.
Read the full abstract

All authors Chinmay Deo, Danish Mahajan, Soudeep Ghoshal, Himanshu Buckchash

Audio and Speech Processing

A Parameter-Free Few-Shot Evaluation for Elephant Vocalisation Classification

Christiaan M. Geldenhuys and Thomas R. Niesler

We present a parameter-free episodic evaluation of nearest-centroid classification of elephant vocalisations on fixed pretrained embeddings, for the Elephant Voices (EV) and Linguistic Data Consortium (LDC) datasets. We ask not which embedding yields the best classifier trained on all labelled data, but how the simplest classifier performs as the number of exemplars per class varies. There are no learnable parameters, because each class is modelled as the mean of its support embeddings and each query is assigned to the nearest centroid under squared Euclidean distance. Evaluation covers the fixed Perch (ver. 1), Perch (ver. 2) and HuBERT (base, layer 2) embeddings, alongside mel frequency cepstral coefficient (MFCC) features, $N$-way $k$-shot, under the same stratified $K$-fold cross-validation protocol as the trained classifiers. None of these embedding models was trained to distinguish elephant call types. On the smaller EV dataset the centroid classifier is markedly data-efficient. Using Perch (ver. 1) or Perch (ver. 2) embeddings it overtakes in mean average precision (mAP) the fully-trained logistic regression (LR) baseline from one or two exemplars and the recurrent baseline from two. Over the reduced set of call types on which the strongly-supervised end-to-end baseline was trained, the centroid classifier using Perch (ver. 2) embeddings overtakes that baseline in mAP as well, from two exemplars. On the larger LDC dataset the recurrent baselines retain their advantage for all considered values of $k$. Only LR is overtaken, and only in mAP. Nearest-centroid classification is therefore preferable precisely when exemplars are few and the fixed embedding already separates the call types.
Read the full abstract

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

What Makes a Good Layer? Assessing the Layer-Wise Intrinsic Properties of Music Foundation Models

Angelos-Nikolaos Kanatas, Yuexuan Kong, Pablo Alonso-Jiménez et al. (+2 more)

Music foundation models are commonly used as frozen audio feature extractors, yet selecting which layer to extract from remains largely heuristic. Current practice defaults to fixed depths or multi-layer fusion, with limited understanding of why certain layers transfer better across downstream tasks or how representation quality varies with depth and pre-training paradigm. We conduct a systematic layer-wise analysis of 12 music foundation models spanning three pre-training paradigms (masked modeling, autoregressive modeling, and contrastive learning), characterizing their hidden representations through intrinsic geometric and transformation-based properties. Correlating label-free representation-quality metrics with layer-wise performance across 15 downstream tasks, we find that several metrics track layer quality for genre classification, emotion recognition, automatic tagging, and beat tracking, albeit with varying strength across tasks and pre-training paradigms. However, all metrics fail on tonal tasks such as key estimation and chord recognition, indicating that no single property serves as a general proxy for representation quality across music information retrieval tasks. To address this gap, we introduce a pitch-transposition equivariance measure that captures properties missed by these standard metrics, providing a consistent indicator of tonal quality across model families. Finally, we show that intrinsic metrics can serve as effective proxies for layer selection, matching or outperforming trainable multi-layer fusion methods, particularly in limited-data settings.
Read the full abstract

All authors Angelos-Nikolaos Kanatas, Yuexuan Kong, Pablo Alonso-Jiménez, Xavier Serra, Dmitry Bogdanov

Audio and Speech Processing

Separate First, Then Associate: A Two-Stage Approach for Real-World Audio-Visual Speech Enhancement

Tongtao Ling and Zhong-Qiu Wang

Audio-visual speech enhancement (AVSE) aims at extracting target speech from multi-speaker mixtures by exploiting visual cues. Although recent studies have reported strong performance on simulated datasets, the performance, however, often drops dramatically when they are applied to real-world audio-visual recordings. To bridge this gap, the Real-World AVSE Challenge held in the ISCSLP 2026 conference calls for participants to design a practical solution for AVSE under real-world conditions, where speaker overlap, acoustic interferences, room reverberation and visual degradations naturally co-exist. In our submission to the challenge, we propose a decoupled separation-then-association approach. It consists of two stages: a separation stage in which a trained, audio-only model (i.e., not using visual cues) is used to separate input multi-speaker mixture to individual speaker signals, followed by an association stage, where an audio-visual CLIP model is used to identify the separated speech signal with the highest similarity with the target speaker's facial video via cross-modal similarity matching. Evaluation results on the challenge dataset show the effectiveness of our proposed approach.
Read the full abstract

Audio and Speech Processing

Singer-Informed Vocal Source Separation for Multi-Singer Music Mixtures

Jocelyn Xu and Minje Kim

Music source separation systems typically extract a single vocal track and do not distinguish between multiple singers. We study singer-informed vocal source separation for multi-singer mixtures. Our framework introduces a short enrollment recording of a target singer to guide separation through a learned embedding. The singer embedding is incorporated using feature concatenation or feature-wise linear modulation (FiLM), enabling the model to focus on the target singer while suppressing interference. We construct a duet dataset based on DAMP-VSEP with quality filtering and non-overlapping enrollment segments. Experiments on solo and duet settings show that while baseline models perform well for single-singer mixtures, the proposed method improves target-singer extraction in multi-singer cases, increasing target-singer SI-SDR from 0.33 dB to 5.58 dB. Fréchet Audio Distance (FAD) further shows improved perceptual quality and better alignment with target audio distributions. Code and checkpoints are available at https://github.com/jocelynxu01/singer-separation-paper.
Read the full abstract

Audio and Speech Processing

Ambisonics Encoding of Room Impulse Responses using a Device-Agnostic Diffusion Model

Eloi Moliner, Christoph Hold, Juan Azcarreta Ortiz et al. (+5 more)

We address the problem of encoding room impulse responses (RIRs) into high-order Ambisonics (HOA) representations from arbitrary and potentially insufficient or incomplete microphone array measurements. This task is fundamentally ill-posed for microphone arrays with limited spatial capture capabilities, such as irregular or sparse arrays, as classical linear methods fail to reconstruct high-order spatial detail. We introduce a diffusion-based generative framework that models the statistical properties of HOA RIRs. This enables device-agnostic encoding from arbitrary microphone arrays, potentially unseen during data measurement. Our approach incorporates a posterior sampling procedure that enforces consistency between the estimated signals and the measurements while plausibly reconstructing spatial information that is unobservable from the limited measurements alone. Experiments on simulated data demonstrate that our method outperforms linear and neural baselines, achieving accurate HOA RIR estimation up to 12th order. A listening test with binaural renderings, including both simulated and measured RIRs, further confirms that the proposed method yields higher perceptual similarity to reference Ambisonics RIRs than all baselines. The flexibility and accuracy of the proposed framework opens new possibilities for scalable acoustics simulations.
Read the full abstract

All authors Eloi Moliner, Christoph Hold, Juan Azcarreta Ortiz, Sebastian Prepelita, Ishwarya Ananthabhotla, Daniel Wong, Sanjeel Parekh, Sanha Lee

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

H2H Music Improv: A Communication Model and Audio-Visual Dataset for Music Improvisation

Aleksandra Teng Ma, Anthony Cammarota, Jiayi Wang et al. (+4 more)

Current real-time AI improvisation systems lack the communication awareness human musicians rely on: rather than treating communication as a foundational algorithm design concern, most systems layer interaction strategies post-hoc onto generative algorithms through explicit controls and predefined modes. This gap persists in part because no formalized, machine-readable communication model with musicians exists. To address this, we study how expert musicians communicate in free (non-idiomatic) improvisation, unconstrained by prior discussion or agreement. Through a collaborative co-design process with expert improvisers, we derive a communication model that (1) captures how free improvisers negotiate musical ideas and enter stable musical spaces, and (2) is formalized as a machine-readable annotation scheme. We further present the H2H (Human-to-Human) Music Improvisation dataset: six hours of audio-visual expert duo improvisations with clean per-player stems and per-player annotations of both their own intentions and their perception of their partner's intentions. To our knowledge, this is the first such dataset for free improvisation. Together, the communication model and the dataset offer a new lens and resource for studying musician communication and may in future inform the design of AI musical partners that communicate by design.
Read the full abstract

All authors Aleksandra Teng Ma, Anthony Cammarota, Jiayi Wang, Alexandria Smith, Cheng-Zhi Anna Huang, Jeffrey Albert, Alexander Lerch

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

The MPB Corpus: A Dataset of Melody, Rhythm, Harmony, and Melody-Harmony Relationships in Brazilian Popular Music

Carlos de L. Almada, Hugo T. de Carvalho and Felipe D. Martins

PLEASE CHECK AND REFER TO THE PUBLISHED VERSION OF THIS PAPER ON https://doi.org/10.5334/tismir.357 This paper presents the MPB Corpus, a dataset of Brazilian Popular Music (Musica Popular Brasileira, MPB) comprising derived symbolic annotations for 500 pieces across four musical parameters – melodic contour, melodic rhythm, harmony, and the melody-harmony relationship – making it the most comprehensive dataset to date for computational musicology of Brazilian music. It is released under a CC-BY-4.0 license in an open, machine-readable tabular format, with documentation and reproducible analysis code. Because the annotations were derived from copyrighted musical scores, the corpus contains no audio recordings or score-level representations such as MIDI or MusicXML; it nonetheless provides a reusable resource for MIR research on a culturally significant repertoire that remains underrepresented in existing datasets. To support the encoding process, we introduce analytical models for capturing rhythmic and melodic information, along with tailored visualizations and summary metrics. Finally, we present a brief exploratory analysis illustrating the dataset's potential for formulating and addressing musicological questions about the genre.
Read the full abstract

Thursday 13 August

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

Machine Unlearning for Speech Question Answering in Large Audio-Language Models

Zhe Liu

Large Audio-Language Models (LALMs) have recently shown strong capabilities in speech understanding and question answering (QA), but they also inherit privacy risks from large-scale training data, including the unintended memorization of sensitive information. In this work, we study machine unlearning for speech QA in LALMs, a setting that is more challenging than prior work on text-based Large Language Models (LLMs) or Automatic Speech Recognition (ASR) due to the tight coupling between acoustic perception and factual knowledge. We present and evaluate multiple unlearning strategies, including gradient ascent, task arithmetic, and alignment-based fine-tuning methods that enforce safe refusal responses, to remove private knowledge while still preserving performance on core capabilities. Through extensive experiments on speech QA datasets, we show that these unlearning methods can reduce the privacy leakage rate by up to 80% while maintaining near-neutral performance on non-private speech QA and general speech understanding benchmarks.
Read the full abstract

Audio and Speech Processing

VoiceChat-TTS: A Low-Latency Continuous Speech Synthesis Model for Interactive Agents

Edresson Casanova, Jaehyeon Kim, Mariana Graterol Fuenmayor et al. (+17 more)

Spoken dialogue is a natural form of human–computer interaction, yet most speech language models remain limited to turn-based operation and lack real-time adaptability, such as user barge-in. Recent duplex speech-to-speech and speech-to-text models reduce latency by replacing multi-stage pipelines, but often compromise speech quality because accurate ASR, interruption handling, and high-fidelity synthesis must be optimized jointly. We propose VoiceChat-TTS, a low-latency, continuous, and streamable text-to-speech model for interactive agents. VoiceChat-TTS is driven directly by LLM text-token streams, supports explicit interruption via control tokens, and produces silence when no textual input is available. The model enables always-on, responsive speech generation while preserving modularity and high speech quality, and it supports mid-utterance interruptions without resetting the KV cache.
Read the full abstract

All authors Edresson Casanova, Jaehyeon Kim, Mariana Graterol Fuenmayor, Shehzeen Hussain, Viacheslav Klimkov, Valentin Mendelev, Mikyas Desta, Paarth Neekhara, Piotr Zelasko, Chen Chen, Elena Rastorgueva, Ke Hu, Ankita Pasad, Xuesong Yang, Aya Alja'fari, Rajarshi Roy, Rohan Badlani, Jason Roche, Jason Li, Zhehuai Chen

Audio and Speech Processing

Trajectory Dynamics in Self-Supervised Learning Latent Space for Audio Deepfake Detection

Tomás Andrade Weber

Human speech production is constrained by physiology, giving rise to characteristic temporal structure on acoustic signals. We hypothesise that these constraints manifest as structured trajectory dynamics in the latent space of Self-Supervised Learning (SSL) models, and that synthetic speech violates them detectably. To test this hypothesis, we train a causal Long Short-Term Memory (LSTM) next-frame predictor on bonafide speech only (Stage 1), using the deepfake-specialised SSL backbone Wav2Vec2-Large-AntiDeepfake, and compare against a static global-average-pooling baseline using identical features, thus isolating the contribution of temporal modelling. A supervised Stage 2, which trains a Multi-Layer Perceptron on the frozen LSTM internal states using labelled data, is included to characterise the role of spoof supervision. Our system achieves competitive or state-of-the-art performance across six benchmarks: ASVspoof 2019/2021, Codecfake, In-the-Wild, MLAAD-EN, and Deepfake-Eval-2024, including best published EER on ASVspoof 2021 (0.75%) and, notably, Stage 1 trained on bonafide speech only surpasses the published supervised baseline from the same backbone on DE2024 (30.35%). On near-domain benchmarks, static and dynamic approaches perform comparably. On harder cross-corpus benchmarks with diverse synthesis methods, trajectory dynamics provide substantial gains, confirming that temporal physiological constraints carry detection signal beyond utterance-level statistics.
Read the full abstract

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

StreamHear: Domain-Adapted Pseudo-Labeling for Semi-Supervised Streaming Speech Recognition

Zefang Liu, Chenyang Zhu, Sangwoo Cho et al. (+3 more)

Streaming automatic speech recognition (ASR) underperforms on domain-shifted target audio, where labeled in-domain data is costly to prepare while unlabeled audio is abundant. We present StreamHear, a semi-supervised pipeline that adapts a pretrained streaming student by fine-tuning an offline transducer teacher on the labeled training set, generating pseudo-labels on the unlabeled portion, and fine-tuning the student on the mixture. We further introduce a prior-regularized dynamic-programming realignment step that fixes chunk-level word placement using an ASR-hypothesis anchor. Across four datasets spanning financial calls, prepared read speech, and phone-quality dialogue, StreamHear consistently outperforms supervised student fine-tuning and narrows the gap to the offline teacher.
Read the full abstract

All authors Zefang Liu, Chenyang Zhu, Sangwoo Cho, Xujun Peng, Shi-Xiong Zhang, Sambit Sahu

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

Motor, Cognitive, or Corpus? What Survives Cross-Lingual Transfer in Speech-Based Parkinsons Disease Detection

Serli Kopar, Sam Gijsen, Abner Hernandez et al. (+2 more)

Self-supervised learning (SSL) speech representations achieve strong performance for Parkinson's disease (PD) detection within individual corpora. However, it remains unclear whether these models capture disease-related characteristics or exploit dataset-specific confounds, particularly since most SSL backbones are pretrained exclusively on healthy speech. To investigate this question, we perform a layer-wise analysis of nine SSL speech backbones using a low-capacity logistic regression probe across three languages. We structure the evaluation as multiple scenarios that progressively introduce distribution shifts in participant identity, recording conditions, language, and pathology. Our results reveal two key findings. First, layer selection is highly corpus-dependent: the optimal representation layer is determined primarily by the source dataset rather than by the SSL architecture itself. Second, the transferred discriminative signal lacks pathological specificity: classifiers trained to detect PD assign similarly high probabilities to both PD and dementia speech in the target corpus. These results highlight critical limitations that must be addressed before speech-based pathology recognition models can be reliably deployed in clinical settings.
Read the full abstract

All authors Serli Kopar, Sam Gijsen, Abner Hernandez, Paula Andrea Perez-Toro, Kerstin Ritter

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

CASA: Content-Acoustic Speaking Assessment with Speech Encoder and Large Language Model

Nhan Phan, Ilona Lähteenmäki, Anna von Zansen et al. (+4 more)

Research on automatic speaking assessment (ASA) has increasingly adopted multimodal speech large language models to assess learners' speaking performance. However, existing studies provide limited analysis of how acoustic and content information contribute to predictions and how stable the resulting performance is. We propose CASA, a simpler architecture combining Whisper-medium and Qwen3.5-2B that achieves state-of-the-art performance while providing a more interpretable separation between speech delivery and content. On the Speak & Improve Corpus 2025, CASA achieves a root mean square error (RMSE) of 0.358, improving on the previous best RMSE while using approximately half the estimated inference parameters. The general-purpose architecture is designed for adaptation to other ASA corpora without structural changes and relies on three handcrafted fluency features. Through ablations and repeated runs, we analyze the individual and complementary contributions of acoustic and content information, examine performance variability, and demonstrate the potential of large language model reasoning for training-free content validation.
Read the full abstract

All authors Nhan Phan, Ilona Lähteenmäki, Anna von Zansen, Olli-Pekka Pauna, Yaroslav Getman, Tamás Grósz, Mikko Kurimo

Wednesday 12 August

Audio and Speech Processing

VoiceDesigner: Text-to-Voice Generation and Editing via Unified Diffusion Modeling and Data Augmentation

Jiarui Hai, Karan Thakkar, Ke Chen et al. (+5 more)

Recent breakthroughs in generative models have made text-to-voice generation (TTV) possible, enabling the synthesis of speech directly from textual voice descriptions. However, existing systems face two key challenges. First, they struggle to generate a diverse range of voices, spanning real-world human speakers and fictional characters. Second, they lack robust and flexible voice editing capabilities, such as voice cloning and the ability to modify attributes like emotion and tone. In this paper, we propose VoiceDesigner, a unified framework for text-to-voice generation and editing that supports diverse and controllable voice design. To tackle the above challenges, we propose solutions from two perspectives. First, we develop a hybrid data pipeline that leverages digital signal processing techniques and speech generation models to construct a diverse voice dataset covering both real-world and fictional voices. Second, we introduce a diffusion transformer with architectural improvements to better handle complex conditioning and enhance multi-task performance, enabling unified voice generation and editing. Through subjective and objective evaluations, VoiceDesigner achieves superior prompt alignment with both voice descriptions and editing instructions, while maintaining competitive perceptual quality and voice usability compared to state-of-the-art TTV models.
Read the full abstract

All authors Jiarui Hai, Karan Thakkar, Ke Chen, Yunyun Wang, Jiaqi Su, Rithesh Kumar, Mounya Elhilali, Zeyu Jin

Audio and Speech Processing

Evaluating Pre-trained Speech Encoders for Spontaneous Speech Detection and Out of Domain Synthetic Speech Generalisation in Indic Languages

Varun Rai, Pavan Kumar J, Sujith Pulikodan et al. (+1 more)

Transformer-based models have shown strong accuracy in distinguishing spontaneous from scripted speech and natural from synthetic speech, but these results are established on a narrow set of well-resourced language benchmarks and have not been extended across Indic languages, nor has embedding geometry been used to explain encoder behaviour or deepfake generalisation failure. We address these gaps by evaluating five frozen transformer encoders, AST, Vaani-FastConformer, Wav2vec2, Whisper and BEATs, across 22 Indic languages, and by conducting a multi-system TTS generalisation experiment across four TTS models. Beyond accuracy, we present language isolation probing and centroid proximity analysis. Probing reveals an encoder-dependent trade-off between language-discriminability and spontaneity detection. Centroid analysis shows that out-of-domain generalisation is predicted by a training system's proximity to unseen TTS embeddings, not its distance from natural speech, a finding with direct implications for training data selection in real-world deepfake detectors.
Read the full abstract

All authors Varun Rai, Pavan Kumar J, Sujith Pulikodan, Nihar Desai

Audio and Speech Processing

Rethinking Language Model-Based Generative Speech Enhancement in the Latent Space of a Neural Audio Codec

Yihui Fu, Zhengyang Li and Tim Fingscheidt

Language model (LM)-based speech enhancement (SE) has recently emerged rapidly using latent space features of neural audio codecs (NACs). In this paper, first, we present a unified framework covering six popular LM-based generative SE modeling paradigms based on discrete/continuous latent NAC features: discrete or continuous autoregressive (D/CAR) SE, discrete or continuous non-autoregressive (D/CNAR) SE, discrete diffusion (DDiff) SE, and continuous flow matching (CFM) SE. Second, we are the first to compare their performance in a unified experimental setup and synopsis with diverse intrusive and non-intrusive metrics, enabling a fair and comprehensive evaluation. Third, we propose a fine-tuning strategy with auxiliary losses on reconstructed speech to improve both intrusive and non-intrusive metrics. Trained and evaluated on URGENT 2025 Speech Enhancement Challenge data splits, all continuous-domain paradigms excel their discrete-domain counterparts. The overall best approach turns out to be CNAR. We further show that our proposed auxiliary loss fine-tuning strategy helps to improve DNSMOS, NISQA, PESQ, and POLQA consistently in all six paradigms.
Read the full abstract

Audio and Speech Processing

The SLT 2026 SmartGlasses Challenge: Benchmarking Egocentric Multi-Talker Speech Recognition and Understanding with Audio-Language Models

Dehui Gao, Zhixian Zhao, Zhennan Lin et al. (+14 more)

Recent advances in large language models (LLMs) and multimodal LLMs (MLLMs) have created new opportunities for wearable speech interfaces, with smart glasses providing an egocentric platform for continuous audio sensing and assistance. However, speech recognition and understanding in this setting remain challenging because of dynamic acoustic conditions, speaker overlap, and the spatial ambiguity introduced by wearer-centered recording geometry. To support systematic evaluation in this setting, we introduce the IEEE SLT 2026 SmartGlasses Challenge for egocentric multi-speaker speech processing. The challenge consists of two tracks, Dyadic Dialogue Understanding and Multi-party Meeting Understanding, and jointly evaluates Time-Stamped Speaker-Attributed Automatic Speech Recognition (TSA-ASR) and Spoken Language Understanding (SLU). It is built on a 106-hour four-channel egocentric speech dataset containing 714 sessions collected in real-world scenarios. This paper describes challenge tasks, dataset construction, submissions, and summarizes the main findings from the shared evaluation. The results show that heavy speaker overlap remains a major factor affecting TSA-ASR performance, while paralinguistic acoustic understanding continues to be difficult for current audio-language models in complex SLU settings. Further details can be found on the official challenge website.
Read the full abstract

All authors Dehui Gao, Zhixian Zhao, Zhennan Lin, Yujie Liao, Yuhang Dai, Yike Zhu, Longshuai Xiao, Hui Bu, Xin Xu, Xie Chen, Shuai Wang, Liumeng Xue, Zhonghua Fu, Jun Du, Eng-Siong Chng, Jun Zhou, Lei Xie