Daily Papers — 2026-09-09
29 papers on audio, speech, music, and acoustics.
1. Do speech foundation models really learn words?
Authors: Robin Huo, Ewan Dunbar
Categories: cs.CL, cs.SD | Proceedings of Interspeech 2026
Self-supervised speech foundation models are now used in a wide array of downstream applications, including traditional speech recognition and as the basis for tokens in speech-aware language models. Attempts to understand their usefulness have largely focused on probing their representations’ ability to discriminate phonemes and words. However, discriminative ability for words need not imply specialized representation of words per se. Good discrimination of words may be explained by good encoding of word form (phonemes) rather than form-independent word representations encoding identity or syntactic/semantic properties. By partialling out phoneme information using residualization, we show that, in later layers, HuBERT and wav2vec 2.0 do in general learn representations which encode words with reasonable fidelity independently of local phonetic content. We show that this simple approach to disentanglement can enhance higher-order linguistic information in word discovery tasks.
2. StreamAlign: Streaming Text-Aligned Speech Tokenization
Authors: Kang-wook Kim, Jinyoung Park, Jinsoo Kim, Sehun Lee, Sang Hoon Woo et al.
Categories: cs.CL, cs.SD, eess.AS | Findings of EMNLP 2026. Project page: https://ishlove77.github.io/StreamAlign/
Text-aligned speech tokenization methods have emerged to better align speech tokens with LLM token spaces, enabling more effective utilization of pretrained LLMs. However, they rely on offline automatic speech recognition (ASR), leading to two key limitations: (i) the need for complete utterances before tokenization, precluding real-time streaming, and (ii) vocabulary mismatch between ASR and LLMs, which reduces acoustic granularity from the subword to the word level. We introduce StreamAlign, a text-aligned speech tokenization framework that enables streaming tokenization for real-time speech-text joint modeling. StreamAlign performs online speech-text alignment by combining character-level RNN-Transducer alignment with word-level ASR guidance, mitigating ASR-LLM vocabulary mismatch while preserving recognition accuracy. A proactive word boundary classifier anticipates word completion at chunk boundaries, reducing tokenization latency from 560 ms to 270 ms. On LibriSpeech, StreamAlign achieves the lowest WER and highest UTMOS among evaluated tokenizers. Furthermore, StreamAlign-SLM, a spoken language model trained on StreamAlign units, outperforms other end-to-end spoken language models in speech continuation while achieving the strongest overall consistency on SALMon and spoken StoryCloze.
3. Who Are They to Each Other? Multi-Agent Reasoning for Speaker Relationship Inference
Authors: Yaohan Guan, Yen-Ju Lu, Yuzhe Wang, Junhyeok Lee, Jesus Villalba et al.
Categories: cs.MA, cs.CL, cs.SD
Inferring speaker relationships from spoken conversations is an important step towards socially aware speech understanding. However, this task remains underexplored, and supervised modeling is costly to train and scale. At the same time, existing inference-time LLM approaches provide limited structure for handling subtle, distributed, and multimodal relational cues that may support multiple plausible interpretations. To address these limitations, we introduce a training-free multi-agent reasoning framework that organizes inference through structured interaction among LLM agents, allowing relationship judgments to be proposed, challenged, and adjudicated without task-specific training. We instantiate this framework with two complementary designs. We propose Multi-Role Multi-Agent Debate as a task-specific adaptation of standard multi-agent debate for speaker relationship inference, assigning agents complementary roles or social-theory-grounded perspectives rather than a single undifferentiated viewpoint. In contrast, we introduce Multi-Agent Compete, a competition-based protocol that compares agent judgments through pairwise adjudication, eliminates weaker candidates, and retains the most defensible one. We evaluate these methods on the Seamless Interaction dataset across different modality settings, covering both binary classification and fine-grained relationship-detail prediction. Results suggest that they improve over zero-shot and existing multi-agent baselines in most cases. Human evaluation further suggests that this task is challenging even for people. LLM methods can sometimes outperform human annotators in text-included settings but are less competitive in the audio setting. Together, these findings suggest that relationship inference benefits from structured inference-time interaction among agents, while acoustic cues are not yet fully captured by current models.
4. Beyond Accuracy: ARIA-Rubrics for Evaluating Audio Reasoning in Large Audio Language Models
Authors: Yupei Li, Qiyang Sun, Mohamed Mady, Chenxi Wang, Zhengwei Gong et al.
Categories: cs.SD | Accepted at AACL 2026
Large Audio Language Models (LALMs) have shown strong performance on audio reasoning benchmarks, but accuracy alone cannot distinguish true reasoning from superficial pattern matching, often overestimating reasoning ability since high scores may result from guessing rather than genuine audio understanding. Evaluating the reasoning process itself is essential for improving LALMs’ reasoning ability, yet remains challenging. Existing methods either rely on costly human annotation or opaque LLM-as-judge approaches, making them impractical, biased, and lacking transparency. Moreover, audio reasoning introduces unique challenges absent in text-based settings, perceptual hallucination and cross-modal alignment between audio understanding and textual inference, hence text-based evaluation frameworks cannot be directly applied. Therefore, we propose ARIA-Rubrics (Audio Reasoning Integrity Assessment), a lightweight, annotation-free gold reasoning chains, automatic and transparent framework comprising six complementary metrics that evaluate audio reasoning quality across perceptual grounding, reasoning coherence, and answer consistency. We use Chain-of-Thought prompting as an externalization mechanism to make the reasoning process observable. Experiments on 9 models across 2 benchmarks identify three reasoning modes of current LALMs with actionable directions for future development, with ARIA-Rubrics achieving high correlation with human judgments. The code is available at the Github Repository.
5. Learned Continuous Synthesis of Quadratic Difference Tone Spectra
Authors: Esteban Gutiérrez, Behzad Haki, Christopher Haworth, Xavier Serra, Rodrigo Cádiz
Categories: cs.SD | Accepted at the AI Music Creativity (AIMC) 2026 Conference held in Berlin, Germany. 8 pages and 3 figures
Quadratic difference tones (QDTs) are a species of auditory distortion product in which a “phantom” pure tone, absent from the acoustic signal, is clearly audible to listeners. Exploiting this phenomenon, one can synthesize harmonically rich tones for musical purposes, a technique called Quadratic Difference Tone Spectrum (QDTS) synthesis. Previous works have introduced numerical methods to synthesize QDTS based on the distortion function, which links a target QDTS and an overtone-structured carrier signal. While accurate, these methods were stochastic and discontinuous, making them difficult to control for musical purposes and effectively limiting them to stationary signals. This paper proposes a neural network-based approach that learns an approximate inverse of the distortion mapping in an autoencoder-like configuration, producing a continuous approximation that addresses prior limitations. Experimental results show that, although slightly less numerically precise, the method is sufficient for perceptual and musical applications. We also implement a real-time version in Max and evaluate its performance. Various sound examples demonstrate its expressive and musical potential. The source code, audio examples, tutorials, and software accompanying this work are available at https://cordutie.github.io/projects/qdts.html
6. Sparse Weight and Edge Circuit Discovery in Transformer-based Acoustic Models
Authors: Jiankun Wei, Ewan Dunbar, Gerald Penn
Categories: cs.SD
Transformer-based foundation models are powerful but opaque, motivating Mechanistic Interpretation methods to uncover the black-box by identifying small computation subgraphs responsible for a task. DiscoGP is a joint weight-and-edge circuit discovery framework originally developed for text decoders. We extend DiscoGP to speech encoders and present, to our knowledge, the first circuit discovery study for modern speech foundation models. Across HuBERT and Wav2Vec 2.0 on several speech classification tasks, we find that the discovered circuits are extremely compact, yet often match or even exceed the performance of the full pretrained encoder with the same downstream head. Through ablations, we show that these circuits reflect pretrained computation rather than random structure or task-head artifacts. We also introduce a memory-efficient DiscoGP variant that reduces the GPU memory cost of edge-circuit discovery at runtime from quartic to cubic. Overall, our results broaden Mechanistic Interpretation beyond text decoders and show that circuit-level analysis can reveal both explanatory structure and unexpected functional behavior in speech encoders.
7. NOPE-HYPE: A Structured Simulation Workflow for Robust Speech-to-Text Across Diverse Acoustic Environments
Authors: Niramay M. Patel, Bibek Behera, Raksha Sharma
Categories: cs.SD, cs.AI, cs.CL, cs.LG
Robust speech-to-text translation systems should perform reliably across diverse acoustic conditions, yet practical pipelines lack controllable tools for systematic environment exploration. Large speech models remain sensitive to unseen acoustic conditions, as training data rarely cover the full range of real environments.We present NOPEHYPE, a structured training workflow that combines a controllable environment simulator, coverage-optimal environment reduction on Power Spectral Density (PSD) templates, and a small, interpretable hyperparameter search over simulator knobs. We show that simulator-generated noise achieves performance comparable to balanced realnoise training across Whisper and SeamlessM4T models, provide principled environment prototype sets, and identify practical default simulator configurations from a structured 27-run hyperparameter sweep.
8. ScorePrompts: Natural-Language Exploration of Symbolic Music Scores through Analysis
Authors: Emmanouil Karystinaios, Gerhard Widmer
Categories: cs.SD, cs.CL, eess.AS | Accepted as a Late Breaking Demo (LBD) at the International Society of Music Information Retrieval conference (ISMIR) 2026
We present ScorePrompts, an interactive system in which users upload a score, receive natural-language descriptions of its musical structure, ask questions about specific passages, and inspect the corresponding analysis results in staff notation. Specialist MIR components first estimate harmony, tonality, cadences, formal boundaries, texture, and note-level roles, organizing their outputs at note, beat, measure, and piece levels. A schema-constrained language model converts these results into descriptions rather than inferring musical structure directly from raw MusicXML. For questions such as “What changes in measures 14-18?”, a deterministic router selects the relevant measures and analytical levels and returns a concise response together with the underlying results and caveats. Verovio renders the score and links the returned information to cited measures and note-level attributes. The interface also exposes intermediate tables and disagreements between analytical levels. ScorePrompts is intended for exploratory score analysis and explanation, not score editing. The demo shows how existing analysis models, constrained language generation, Q&A retrieval, and notation-based visualization can provide natural-language access to symbolic music analysis while keeping intermediate results inspectable.
9. TimeCues Studio: A Workspace for Music Annotation and Algorithm Prototyping
Authors: Sapir Caduri, Yoav Goldberg
Categories: cs.SD, cs.HC, cs.LG, cs.MM | 8 pages, 2 figures, to appear in Proceedings of the 34th ACM International Conference on Multimedia (MM ‘26)
Multimedia applications require precise music annotation-labeled positions, segments, or loops-placed by hand or algorithmically. Machine-learning algorithms are scalable and effective but need annotated training data, scarce for many tasks. TimeCues Studio is an open-source workspace where algorithm-development teams annotate a music corpus, compare detection algorithms against those annotations, and prototype new ones. Unlike existing tools built for a single track at a time, TimeCues targets teams annotating whole collections, tightly integrated with algorithm development. Annotators place several marker types-each supporting ambiguity-aware labeling-on a grid-locked timeline that visualizes many music features, including separated audio stems. The same timeline drives an algorithm-comparison engine with bundled baselines, a Python sandbox for prototyping new models, and an ambiguity-aware evaluator that honors the structured fields. The same visualization suits solo annotators on music-sync projects. TimeCues is MIT-licensed and deploys via one Docker Compose command.
10. Project Qualia: Recovering Experiential Music Structure from Session Co-occurrence Data
Authors: Nizam Mohammed, Abu B. S. Rahman, Dimuthu D. K. Arachchige
Categories: cs.SD, cs.IR, cs.LG | 10 pages, 3 figures, 4 tables
This report presents results from Project Qualia, an ongoing effort to determine whether experiential similarity between songs, a structure not captured by genre or metadata taxonomies, can be recovered from real listening behavior. We constructed a large-scale dataset of listening sessions, comprising 1.29 billion scrobbles collected from 9,396 users via the Last.fm API and reduced through a preprocessing pipeline to 531.6 million training scrobbles across 28.6 million sessions. On this corpus, we trained a skip-gram Word2Vec model (Song2Vec), treating each session as a sentence and each track as a token. As anticipated, the resulting embedding space was dominated by artist identity, a consequence of single-artist runs within sessions. To test for a subtler, artist-independent signal, we developed an artist-residual procedure: subtracting each artist’s centroid from its tracks’ embeddings and evaluating whether the remainder retained structure. Mean cross-artist cosine similarity fell from 0.2487 in raw embedding space to 0.0005 in residual space, yet 4,577 cross-artist track pairs retained cosine similarity $\ge 0.70$ in residual space, forming coherent genre- and era-based clusters, including trip-hop, 1990s grunge, 2020 mainstream pop, and cross-composer classical piano pairs at cosine similarity up to 0.95. These results confirm that the training data contains experiential structure independent of artist identity, establishing an empirical basis for an architecture designed to learn this experiential layer directly.
11. Orukeet: Multilingual ASR with Frozen Gabor Kernels
Authors: Nathan Roll, Irene Yi, Büşra Marşan, Vianney Grenez, Gabriel Stein et al.
Categories: cs.SD, cs.LG, eess.AS | 5 pages, 2 figures. Code and model: https://github.com/Oruk-AI/orukeet
Orukeet replaces half of an adapted Parakeet encoder’s temporal filters with 12,288 fitted Gabor kernels, freezes these replacements, and trains the remaining parameters on multilingual and multi-accent data. Final adaptation and checkpoint selection use LibriSpeech test-other. Across 20,146 FLEURS recordings in 25 languages, pooled word error rate (WER) falls from Parakeet’s 11.01% to Orukeet’s 9.85%, a 10.6% relative reduction. Orukeet has lower WER on 23 of the 25 languages. Orukeet outperforms Parakeet on 61 out of 74 tested splits, including LibriSpeech test-clean (1.46% vs. 1.53% WER), test-other (2.86% vs. 3.14%), and FLEURS English (3.82% vs. 4.28%). All comparisons decode the same audio with matched NeMo settings. The fitted kernels are stored as ordinary convolution weights, retaining Parakeet’s architecture and inference operators.
12. Unifying Score and Performance for Fine-Grained Music Understanding in Audio-Language Models
Authors: Milan Liessens Dujardin, Song-Ze Yu, Kevin Miao
Categories: cs.SD, eess.AS
Large audio language models (LALMs) have shown promising progress in broad music-understanding tasks such as tagging, retrieval, and captioning. Music understanding that requires finer hearing over both the content and how it is realized within a performance through dynamics, phrasing, articulation, time, and other performance techniques, however, remains at an earlier stage. Existing audio-language model (ALM) training pipelines typically rely on coarse, weakly grounded captions and therefore provide little support for learning these subtle nuances in music, limiting their ability to serve real-world applications in education or artistic practice. We therefore introduce MuNo-SP (Music Notation unifying Score and Performance), a text-based representation that jointly encodes score content and performance information. Building on MuNo-SP, we develop an automatic training-data generation pipeline that uses aligned scores and performances to produce long-form auditory analyses and musically informed question-answer pairs. We use this pipeline to construct MAESTROCaps, a classical piano dataset comprising 148 long-form performance analyses and 31,080 question-answer pairs derived from 148 aligned score-performance pairs. In a human evaluation, MuNo-SP analyses were preferred by majority vote over MIDI-only analyses for eight of nine excerpts. MuNo-SP also performed strongly on a benchmark of score-performance understanding, suggesting that integrating score and performance information enables more reliable and musically informative LALM supervision than a MIDI-only baseline.
13. From Metrics to Natural Dialogue: French Full-Duplex Benchmark for Spoken Dialogue Models
Authors: Hamid Soltani, Gilles Boulianne
Categories: eess.AS | 9 pages, 0 figures. Accepted for publication in Proceedings of the Conference on Empirical Methods in Natural Language Processing (EMNLP): Industry Track, 2026
Full-duplex spoken dialogue models aim to make voice agents more natural by allowing them to listen, speak, pause, and respond during ongoing conversation. However, it is not clear whether full-duplex benchmarks behave the same way when models are evaluated in a different language. To investigate this, we introduce a French full-duplex benchmark (FDB) with two variants, CALLFC-FDB for Canadian French and MEDIA-FDB for European French, and compare them with an English FDB. Built from real spoken resources, these benchmarks evaluate key full-duplex skills, including pause handling, turn-taking, backchannels, and interruptions. Beyond introducing French FDBs, we evaluate human–human conversations with FDB metrics to better understand the values these metrics take in real-world dialogue. Our analysis reveals that most timing-based metrics behave similarly across languages, while content-based evaluation degrades under language mismatch. We also find a trade-off between optimizing benchmark metrics and preserving conversational naturalness.
14. Phoneme-Aware Pronunciation Representations for L2-English L1-Background Accent Identification
Authors: Yangyang Qu, Massimiliano Todisco, Nicholas Evans
Categories: eess.AS
We study speaker-disjoint accent identification for L2 English, where the goal is to predict a speaker’s first-language (L1) background from English pronunciation. Most existing systems classify accents using a single utterance-level representation, but such global representations can obscure pronunciation cues that depend on specific English phonemes. We propose a transcript-assisted model that makes phoneme information explicit during accent identification. Instead of representing an utterance only as a global speech embedding, we represent it as a sequence of pronunciation units, each combining acoustic evidence from a spoken segment with the aligned English phoneme for that segment. A frozen speech encoder provides the acoustic features, while the transcript is used only to obtain phoneme-level forced alignments. No word-level or sentence-level text representation is passed to the accent classifier. Under a four-fold speaker-disjoint protocol on L2-ARCTIC, our model achieves 81.41% accuracy and 81.21% macro-F1, the highest mean performance among the evaluated systems. Diagnostic ablations support the importance of phoneme-aligned token construction, while a Whisper-based ablation shows an additional gain from phoneme information.
15. SCNet: Enhancing GAN-based Speech Generation with Subband Condition Network and Magnitude-aware Phase Loss
Authors: Nan Xu, Mingxue Yang
Categories: eess.AS | Accepted by Interspeech 2026
Recent speech generation has been predominantly driven by GAN-based networks aimed at high-quality waveform synthesis from mel-spectrograms. However, these methods often operate as black-box models, leading to the loss of inherent spectral information. In this work, we propose SCNet, a GAN-based vocoder augmented with a Subband Condition Network to address this issue. Specifically, SCNet leverages a subband signal predicted by a lightweight condition network as prior knowledge. This subband signal is then transformed via STFT to obtain Fourier coefficients, which are integrated into the backbone for the enhanced reconstruction. Additionally, to mitigate the phase wrapping, we introduce a magnitude-aware phase loss that computes instantaneous phase errors weighted by the corresponding magnitude, emphasizing regions with higher energy. Experimental results demonstrate that SCNet achieves superior performance in both objective and subjective evaluations for high-quality speech generation.
16. SphereVAE: Hyperspherical Latent Autoencoders for Robust Autoregressive Speech Representation Modeling
Authors: Haoyu Zhang, Jingbin Hu, Hanke Xie, Qirui Zhan, Wenhao Li et al.
Categories: eess.AS | 15 pages, 4 figures. Accepted to NCMMSC 2026
With the rapid development of speech generation technology, discrete codec representations have been widely used because they provide a stable prediction paradigm. In expressive speech generation, however, the quantization bottleneck of discrete codecs results in information gaps in fine-grained prosody, timbre, pronunciation, and frame-to-frame continuity. Continuous representations (e.g., VAE latents), by eliminating this constraint, have emerged as a more effective alternative for autoregressive modeling. Yet when continuous representations are used as autoregressive prediction targets, prediction errors can accumulate along the generation chain, causing latent drift and degrading long-form stability. To mitigate this problem, we propose SphereVAE, which constrains the VAE latent space to the unit hypersphere. SphereVAE defines a Power Spherical posterior on the hypersphere and regularizes the latent distribution toward a uniform prior, so that information is encoded mainly by directional variation, providing a bounded geometric target for autoregressive prediction and reducing the risk of norm drift. SphereVAE underperforms the standard VAE on reconstruction metrics due to reduced latent freedom. However, when integrated into VoxCPM for zero-shot TTS and long-text generation, it yields lower content error rates with comparable speaker similarity, and shows more stable long-range speaker consistency. These results indicate that an appropriate latent geometric constraint can effectively mitigate autoregressive error accumulation and drift in speech generation.
17. Teacher-Free Self-Distilled Consistency Trajectory Learning for Fast Speech Enhancement
Authors: Shuubham Ojha, Carol Espy-Wilson
Categories: eess.AS
Consistency trajectory models offer a route to fast, high-quality speech enhancement, collapsing the many reverse steps of diffusion-based enhancers into a handful. When instantiated on a Schrödinger bridge (SB), which pins the generative process to fixed clean and noisy endpoints, existing consistency-trajectory enhancers (SBCTMs) still require a pretrained teacher to supply trajectory supervision, which raises training cost and ties the final quality to that of the teacher. We propose a teacher-free, self-distilled consistency-trajectory framework that removes the external teacher: trajectory targets are generated by an exponential-moving-average (EMA) copy of the student, and the model is trained with a three-stage curriculum of $\x_0$ prediction, a self-distilled shortcut objective, and perceptual fine-tuning with a multi-resolution short-time Fourier transform (MR-STFT) loss. Using the same NCSN++ backbone as SBCTM, our model attains a wide-band PESQ of $3.01$, ESTOI $0.87$, and SI-SDR $19.07$\,dB on VoiceBank+DEMAND without a teacher. Varying step count and inference schedule we find that a geometric schedule at low reverse step count maximizes perceptual quality, while a higher-step uniform schedule favors signal fidelity, with the geometric advantage narrowing with reverse step count.
18. UniStream: Multi-Expert Residual Vector Quantization for 48 kHz Causal Streaming Audio Coding
Authors: Mingyu Zhao, Zhiyong Wu
Categories: eess.AS | 15 pages, 1 figure, 5 tables. Accepted at NCMMSC 2026
We present UniStream, a fully causal 48 kHz neural audio codec for streaming speech, music, and environmental sounds. At its core is Multi-Expert Residual Vector Quantization (ME-RVQ), which replaces the single shared codebook in each residual quantization layer with four expert codebooks controlled by a deterministic Top-K router. Because routing decisions are derived solely from previously decoded quantized states, the decoder can reproduce the selected experts without transmitting expert identifiers, thereby expanding quantization capacity while adding 5.5M parameters. We further introduce an auxiliary Optimal Transport Conditional Flow Matching (OT-CFM) objective to regularize the quantized latent space during training. The flow module is removed entirely at inference and therefore incurs no runtime overhead. UniStream supports a 12 kbps Top-1 mode and a 22.5 kbps Top-2 mode within a causal 48 kHz encoder-decoder framework, while achieving real-time GPU inference. To complement narrow-band speech metrics, we report 48 kHz ViSQOL audio mode, ViSQOL speech mode, standard VGGish-FAD, DNSMOS P.835, and higher-rate reference comparisons with Opus and EnCodec. At 12 kbps, UniStream-Top1 achieves PESQ and UTMOS scores comparable to EnCodec while reducing speech Mel-D from 13.07 to 8.21. At 22.5 kbps, UniStream-Top2 achieves a ViSQOL speech-mode score of 4.67 and an environmental audio-mode score of 3.96, exceeding all evaluated systems operating at 12 kbps or below in the latter setting. It also comes within 0.03 MOS-LQO of Opus at 24 kbps on speech in ViSQOL audio mode. Ablation studies confirm that ME-RVQ is the primary source of quality improvement, whereas OT-CFM provides perceptual gains on speech with a mild trade-off in spectral distortion.
19. AVSRBench: A Multi-Condition AVSR Benchmark
Authors: Rishabh Jain, Naomi Harte
Categories: eess.AS, cs.CV, cs.MM | Accepted to IEEE SLT 2026
While AVSR has achieved sub-1% word error rates on the standard LRS3 benchmark, its reliance on broadcast speech obscures whether this reflects true generalization or just domain adaptation. To investigate this gap, we evaluate three AVSR architectures across six conditions: controlled broadcast speech, fixed-grammar utterances, hyper-articulated Lombard speech, read speech from professional lipspeakers and non-professional speakers, and spontaneous multi-party video conversations. We find that visual-only performance deteriorates rapidly beyond broadcast domains, and audio-video fusion mainly benefits Lombard speech environments. Visual understanding degrades sharply at 90° profile views, with multimodal systems relying largely on acoustic fallback. Additionally, speaker articulation proves more critical than minor camera shifts, and LLM-based architectures suffer from poor out-of-domain generalization. Our work highlights a significant generalization gap in current AVSR research. To address this, we also introduce RoomReader-AV as a new benchmark for AVSR and release a unified data preprocessing pipeline to make comprehensive multi-condition evaluation accessible.
20. Candor-LR: A Dyadic Conversational Dataset for Audio-Visual Speech Recognition
Authors: Rishabh Jain, Aristeidis Papadopoulos, Zhaofeng Lin, Naomi Harte
Categories: eess.AS, cs.CV, cs.MM | Accepted to IEEE SLT 2026
Current audio-visual speech recognition (AVSR) benchmarks, like LRS3, rely heavily on clean, scripted and rehearsed speech. They fail to reflect the complexity of natural conversation, which involves overlapping speech, spontaneous turn-taking, unscripted vocabulary and variable acoustic conditions. To shift the field toward realistic dialogue, we introduce Candor-LR, a conversational benchmark derived from the CANDOR corpus of 1,656 natural dyadic videoconferences. Our custom data preparation pipeline yields 713.5, 10.1, and 60.1 hours of training, validation, and test data, respectively. Evaluating pretrained AVSR models on Candor-LR reveals that audio-only accuracy drops sharply compared to LRS3, but visual cues compensate effectively, driving much larger performance gains on Candor-LR than on LRS3. Furthermore, training on this corpus significantly improves cross-domain robustness under both clean and noisy conditions, as its realistic conversational data captures broader audio-video features. We open-source our pipeline to ensure reproducibility, establishing Candor-LR as a challenging benchmark for conversational AVSR.
21. Over-Tightening-Aware Pseudo-Labeling for Tight-Boundary Speaker Diarization
Authors: Shota Horiguchi, Takanori Ashihara, Marc Delcroix, Naohiro Tawara, Alexis Plaquet
Categories: eess.AS, cs.SD | Accepted to IEEE SLT 2026
Training speaker diarization models on loose labels, such as speech segments with padded boundaries or filled pauses, often results in similarly loose model outputs. To obtain tighter boundaries, pseudo-labeling based on the averaged outputs of causal and anticausal models has been proposed. However, since the pseudo-labels are estimation-based, they can suffer from over-tightening, which increases missed detections that can propagate as unrecoverable errors to downstream tasks. This paper carefully analyzes the causes of over-tightening and proposes three approaches to address them: (i) removing pause filling rather than padding, (ii) introducing a burn-in phase to mitigate missed detections near the beginning of causal and anticausal predictions, and (iii) making pseudo-label-based co-training aware of the non-causal model used for final inference. Experimental results show that the proposed method reduces missed detections caused by over-tightening and improves both diarization accuracy and downstream multi-talker ASR performance.
22. Pushing the Boundaries of Streaming Multi-Speaker ASR: A Systematic Study of Architectural Trade-offs
Authors: Taejin Park, Ivan Medennikov, Kunal Dhawan, Weiqing Wang, Jagadeesh Balam et al.
Categories: eess.AS, cs.SD | This article is accepted to Interspeech 2026
Streaming multi-speaker ASR is a challenging task that must balance accuracy, latency, and efficiency while handling overlapping speech and maintaining coherent long-context modeling over extended conversations in an online fashion. We present a unified framework that categorizes streaming multi-speaker ASR into four architectural strategies based on how diarization and ASR are integrated. Using a shared pair of open-source streaming ASR and diarization models as a common foundation, we derive four multi-speaker ASR systems that differ in whether they employ multiple model instances, fine-tuning, or both. We evaluate these systems across multi-speaker accuracy, single-speaker accuracy degradation, memory footprint, and training complexity. Through this systematic architectural analysis, we clarify the design space for streaming multi-speaker ASR and provide practical guidance for selecting the most suitable approach under diverse deployment constraints.
23. SpeechAnnotator: A Context-Aware Multi-Agent Framework and Benchmark for Multidimensional Speech Annotation
Authors: Qirui Zhan, Shuiyuan Wang, Jingbin Hu, Haoyu Zhang, Xiaming Ren et al.
Categories: eess.AS, cs.SD | 15 pages, 3 figures, to be published in NCMMSC 2026
Recent controllable speech generation requires training data with fine-grained annotations of speaker traits, prosody, emotion, paralinguistic cues, acoustic scenes, and context. Existing workflows often rely on manual correction, paid hosted multimodal services, or fixed processing chains, which limits large-scale data processing through annotation cost, external-service dependence, or weak cross-stage recovery. We introduce SpeechAnnotator, a locally deployable, context-aware multi-agent framework built entirely from open-source models and tools. Supporting frontend modules first obtain speaker-aware segments and final segment transcripts, while prior evidence extractors attach heterogeneous segment-level cues. Three specialist agents then collaborate through shared state: the Planning Agent converts local audio evidence, speaker history, neighboring segments, and recording-level context into field-specific contracts; the Labeling Agent performs contract-guided multimodal prediction for directly observable attributes; and the Review Agent runs a bounded review loop that checks evidence support and cross-segment consistency, triggering relabeling only for unsupported or inconsistent fields. To address the fragmentation of existing evaluation resources across isolated tasks and narrow-domain test sets, we introduce SpeechAnnotator-Bench (SA-Bench), containing 8.87 hours of human-annotated audio across nine source formats, together with SpeechAnnotator-Eval (SA-Eval), which separates Timeline-Eval for speaker-aware timeline recovery, Closed-Eval for finite-set attributes, and Open-Eval for open-ended attributes. Experiments and ablations show that SpeechAnnotator provides a locally deployable alternative to commercial audio-capable systems, while the bounded review loop improves multidimensional annotation through evidence- and context-aware field-level recovery.
24. Why Learning Rediscovers the Closed-Form Diagonal Regularizer
Authors: Jeahn Han, Pyojin Kim
Categories: stat.ML, cs.LG, cs.RO, eess.AS | main paper: 9 pages, 3 figures appendix
We identify a diagonal saturation principle in modal inverse problems: when truncation noise is isotropic, the Bayes-optimal Tikhonov shape is a closed-form power law Gamma_k proportional to lambda_k^ s set by the prior alone, independent of the domain. Berry’s random-wave conjecture decorrelates the truncation noise across modes, and Weyl’s eigenvalue counting law supplies enough modes for the conclusion to survive empirical Berry violations. Together they predict an approximately flat loss landscape across the per-mode family, leaving narrow scope for a diagonal regularizer to robustly beat the closed form. On FEM-simulated acoustic rooms, the closed form is near-optimal relative to per-room oracle tuning across observation windows, and three diagonal architectures trained on the same data match its reconstruction error within 1 pp despite learning qualitatively different spectra. The framework extends to heat diffusion via a known exponential Green’s function correction with no new free parameters. Saturation is restricted to the diagonal family: Learned Iterative Ridge crosses the boundary by exploiting cross-mode coupling, locating where learning starts to help.
25. $S^3$-Bench: Evaluating Speech Interaction Models as Scientific Voice Assistants
Authors: Heyang Liu, Jiayi Huang, Wenyang Xiao, Ziyang Cheng, Lixin Zhang et al.
Categories: cs.CL
The advance of multimodal large language models (MLLMs) has fundamentally reshaped the paradigm of human-computer interaction, especially speech interaction models capable of seamless conversations. Despite remarkable performance as general voice assistants, their performance in specialized domains remains underexplored, particularly in scientific areas. Scientific interactions introduce formidable challenges, involving rare technical terminology, spoken norms of abbreviations, and the natural verbalization of symbolic special expressions. In this paper, we introduce S$^3$-Bench, a systematic evaluation framework covering 10 major disciplines, consisting of a Knowledge set for speech question-answering and a Dialogue set for multi-turn progressive interactions with simulated user agents. By decomposing a complete atomic turn into stages of speech recognition, perception, knowledge utilization with reasoning, and response pronunciation, we systematically characterize the common challenges and performance tradeoffs of existing approaches. Furthermore, experiments on multi-turn interactions reveal persistent limitations in user adaptation and the generation of accurate, comprehensive, and efficient responses.
26. BuzzASR: A Swarm of 100+ Monolingual Speech Recognition Models
Authors: Shivam Singh, Aditya Yadavalli, Catherine Arnett, Alex Warstadt
Categories: cs.CL | Accepted at EMNLP 2026. Models: https://huggingface.co/BuzzASR ; Project page: https://lemn-lab.github.io/buzz-asr
We introduce BuzzASR, a collection of language-specialized fine-tuned Whisper models adapted for automatic speech recognition (ASR) in 102 languages. Large end-to-end Transformer-based ASR models such as Whisper have revolutionized ASR, but most prominent models are highly multilingual. As a result, these models often perform poorly on languages less well-represented in their training set. While it has long been known that effective language adaptation can be achieved through simple fine-tuning on monolingual data, this strategy has only been applied to a small number of languages. We massively scale up this simple approach to 102 languages covered in the FLEURS dataset, while also implementing a more complex language adaptation strategy that integrates monolingual tokenizer replacement and data augmentation using text-only fine-tuning. BuzzASR models outperform Whisper-large-v3 on 77 out of 102 languages, reducing character error rates (CER) by a factor of over 2.8 on average. Our models achieve state-of-the-art CER among open-source systems on 27 of 102 languages on the combined FLEURS and Common Voice test set. Our tokenizer replacement strategy yields an average 3.3x improvement in compression rate (characters per token) over Whisper’s multilingual BPE, with gains of up to 21.7x. We release all models, code, and detailed results: https://lemn-lab.github.io/buzz-asr
27. Leveraging Fine-grained Error Correction in Korean Speech Recognition for Consultation Services
Authors: Yonghyun Jun, Jimin Lee, Hwan Chang, Dongho Shin, Seolah Kim et al.
Categories: cs.CL | Published in Engineering Applications of Artificial Intelligence
Automatic Speech Recognition (ASR) technology is fundamental to customer service automation and large-scale transcription. However, even advanced ASR models exhibit inevitable errors in complex real-world environments such as call center conversations. When privacy restrictions preclude audio access, error correction must rely on text-based post-editing. Existing text-only approaches face significant challenges in low-resource languages, mainly due to a critical scarcity of annotated corpora and tailored correction methodologies. For Korean, this resource gap is particularly pronounced, as existing resources are predominantly designed for ASR training rather than text-based error correction. To address this, we introduce DasanCallDial, the first large-scale Korean benchmark dataset specifically curated for dialogue-level ASR error correction. Derived from genuine call center interactions, it comprises 1,974 dialogues with 115,460 utterances. Leveraging this resource, we propose Detector-Gated Contextual Span Correction (DCSC), a text-only post-editing framework for error-sparse Korean speech recognition transcripts. DCSC combines an encoder-based detector that first performs token-level error detection, followed by a language model-based corrector trained to rectify fine-grained span-level errors. Additionally, we employ dialogue-level context augmentation to enable the model to leverage discourse history for disambiguation. By employing multi-level granularity, our method achieves state-of-the-art performance, effectively overcoming the limitations of general LLMs in low-resource settings.
28. X2-NativeCursor: Native-Token Text Progress Tracking for Incremental-Text Streaming Codec TTS
Authors: Zehan Liu, Carl Chen, Rime Wen, Kaiqi Fu, Altman Lin et al.
Categories: cs.CL
Incremental-text streaming text-to-speech (TTS) needs online text progress tracking for synchronized highlighting, interruption handling, and dialogue-history updates. Input text arrives before it is spoken, so text arrival alone cannot indicate speech progress. Existing waveform-based alignment requires complete audio or adds acoustic processing during streaming. We propose X2-NativeCursor, a lightweight observer that tracks progress from native speech tokens before waveform decoding without changing the TTS generator. Its normalization plan links spoken labels to their original-text spans. Text and native-token encoders feed a local matcher that estimates the current label position. A separate output rule converts revisable position estimates into a cursor that never moves backward. Mean absolute error against an automatic reference is 0.151 Chinese characters with 80-ms lookahead, versus 1.253 characters with 320-ms lookahead for an online waveform baseline. Alignment real-time factor also decreases from 0.3598 to 0.0180 relative to this baseline. Lower tracking error is retained under a second automatic alignment reference. We evaluate X2-NativeCursor on Qwen3-TTS and validate its adaptation to CosyVoice2 by training a separate observer for each backbone. Code is publicly available at https://github.com/X-Square-Robot/X2Streaming-TTS.
29. The Semantic Bottleneck: Leveraging Semantic Representations for Non-Invasive Speech Decoding
Authors: Gilad D. Landau, Dulhan Jayalath, Oiwi Parker Jones
Categories: cs.CL, cs.LG | 12 pages, 8 figures
Non-invasive speech decoding remains constrained by the low signal-to-noise ratio of neural recordings, which makes fine-grained reconstruction of phonemes or individual words difficult. Motivated by neuroscientific evidence that high-level semantic representations are distributed across cortical regions and evolve over slower temporal scales, we hypothesize that semantic content may provide a more suitable target for non-invasive decoding than low-level acoustic or lexical features. We introduce Brain2Semantics2Text, a method that reconstructs text through an intermediate semantic embedding space. Our model maps sentence-level MEG responses into a semantic manifold and then inverts the predicted embeddings into natural language. This semantic bottleneck enables recovery of high-level meaning without word-level alignment. We describe the core principles of the approach, its implementation, and the strategies used to mitigate the challenges of learning a reliable neural-to-semantic mapping. Finally, we compare against prior non-invasive Brain2Text methods and show improved sentence-level results.