Papers

This paper presents the NPU-ASLP music restoration system developed for Track 1 of the China Computer Federation (CCF) Advanced Audio Technology Competition (AATC) 2026 Challenge. The task aims to recover clean music from mobile-device-captured recordings affected by coupled degradations from playback devices, acoustic propagation, environmental noise and recording terminals. System development is further constrained by limited and imperfectly aligned real data pairs and a strict model-complexity budget. In our system, we adopt TF-MoE as the backbone for its dual-axis modeling of cross-band and temporal dependencies, whose Top-1 routing activates a single expert per token to bound computation under the strict budget. We extend it with two task-specific adaptations: a scale-aware encoder and a complex mapping decoder. The former encodes normalized complex spectra jointly with per-band log-RMS to preserve the band-energy cues discarded by band-wise normalization. The latter directly estimates the clean complex spectrum rather than masking the input, thereby repairing coupled degradations. To address the limited, imperfectly aligned real recordings, we further build a challenge-calibrated playback–propagation–recording simulator that generates 1,572.12 h of sample-aligned data across 202 acoustic conditions. The final system contains 2.951M parameters and requires 82.659 GMACs for a 10-s input. On the official blind test, it achieved -49.067 dB SI-SNR, 1.273 LSD, 2.540 FAD, and 3.495 ViSQOL. It ranked third in LSD, FAD, and ViSQOL and received the third prize in the challenge.

Neural audio codecs achieve remarkable performance but operate as closed ecosystems with rigidly co-trained encoderdecoder pairs, severely hindering interoperability. To resolve this, we propose BridgeCodec, a versatile framework that decouples these pairs by translating latent representations between mismatched, frozen endpoints. We formulate this cross-codec translation as a Schrödinger Bridge optimal transport problem, enabling robust mapping between disparate distributions without predefined noise priors. To efficiently capture longrange speech temporal dependencies, we integrate a Mambaenhanced U-Net architecture. We validate BridgeCodec on an extreme deployment scenario: mapping a lightweight 8 kHz source encoder directly to a 48 kHz target decoder at an ultralow bitrate of 1 kbps. Experimental results demonstrate that BridgeCodec achieves superior wideband reconstruction and high perceptual quality, successfully bridging heavily bottlenecked, heterogeneous codecs.

Speaker anonymization (SA) aims to conceal speaker identity while preserving linguistic content and emotion, but existing streaming systems often lose emotion information through discrete content bottlenecks or require emotion-specific modeling. We present LiveAnon, a voice-conversion-based streaming SA system that directly conditions an autoregressive codec language model on hidden features from a pretrained streaming ASR encoder, preserving emotion without emotion-specific supervision. To reduce codec-generation latency, we introduce capped RVQ decoding, which preserves dependencies among coarse codebooks while predicting the remaining codebooks in parallel. Under VPC24, LiveAnon improves intelligibility and emotion preservation over fixed-target Stream-Voice-Anon (SVA), a strong prior streaming SA system, matches emotion-prompted SVA in emotion preservation, and achieves the strongest semi-informed privacy among streaming systems. Under VPC26, it ranks second across all metrics among the evaluated anonymization systems, demonstrating a strong privacy-utility balance.

Beat tracking, chord estimation, and structure analysis are staple tasks of music information retrieval. Their datasets are drawn largely from Western popular music, and jazz remains a repertoire on which state-of-the-art systems underperform. We present JSIR, the Jazz Standards dataset for Information Retrieval. Its 267 recordings of 44 jazz standards span 25.6 hours, 3.7 times the audio of JAAH, the closest existing resource. For each recording, we provide annotations for beat and downbeat positions, chords, and sections, all produced by a jazz musician and revised in several passes. In baseline experiments, models trained with JSIR also improve on other jazz datasets: the downbeat F-measure of a state-of-the-art beat tracker rises from 87.7% to 89.2%, and the seventh-chord accuracy of a pop-trained chord estimator on JAAH from 34% to 41%. We publicly release the annotations, the annotation tool used to create them, and a web application that displays them in sync with the audio.

As speech generation and real-time spoken dialogue systems advance, neural mean opinion score (MOS) predictors are increasingly used to evaluate speech quality. However, many existing predictors are trained primarily on English read speech and generalize inconsistently across languages and speaking styles. In particular, their reliability on Korean speech has not been adequately validated. We propose KOMOS, a Korean MOS predictor for evaluating the overall quality of both read and spontaneous speech. To train and evaluate KOMOS, we collect 5,000 ratings from 29 native Korean listeners for 1,000 utterances spanning real recordings, synthesized speech, and degraded speech. KOMOS achieves an utterance-level Spearman rank correlation (SRCC) of 0.900, compared with 0.703 for UTMOS v1 and 0.767 for UTMOS v2. The inference code is publicly available at https://github.com/upskyy/komos and the trained model at https://huggingface.co/upskyy/komos.

Removing audio can shift a visual predictor's output without showing whether the task uses the intended audio information. We introduce Conditional Input and Visual Interpretation Checks (CIViC), a four-gate protocol that links an interpretation to an intervention, held-out task comparison, or endpoint control. We demonstrate CIViC on actor-disjoint face-raster tests, a synthetic audio-feature mixture, and source-video-disjoint SpeakerVid proxy tasks. In the static face-raster test, cross-actor audio replacement meets the registered detection rule with a small effect; on a separate temporal endpoint, constant-audio fits score better than audio-trained fits in four matched seeds. In the synthetic mixture, the condition-augmented fit has higher 1,000-step RMSE than the audio-only fit (0.1807 vs. 0.1707), with the gap shrinking during training. In the SpeakerVid proxy, the five-feature-zero fit (five low-level audio summaries zeroed; identity, dialogue, and condition inputs retained) has lower whole-frame grayscale RMSE than aligned audio across four held-out conditions and five paired seeds, while aligned audio has lower segment-drift error. The results show that output activation, held-out error, and endpoint controls support different interpretations across these proxy tasks.

High-fidelity audio coding requires faithful reconstruction of perceptually important high-frequency components, which are often lost due to the aggressive temporal decimation and quantization of neural codecs. State-of-the-art neural audio coding methods rely on adversarial learning to hallucinate these missing components, yielding perceptually plausible but imprecise reconstructions; a limitation for high-fidelity music applications. As a remedy, we propose high-efficiency neural audio coding (HE-NAC), which provides a multi-band extension of the popular residual vector quantization (RVQ) scheme. HE-NAC introduces an effective training mechanism that harmonizes a bitrate-free blind upsampling process, residual learning to compensate it with minimal bitrate, and a multi-head decoder architecture for band-specific reconstruction. As a result, HE-NAC achieves a balanced trade-off, reliably reconstructing the core-band ($\leq$ 3 kHz) component while preserving overall sound fidelity. Subjective listening tests and objective evaluations demonstrate that HE-NAC outperforms both the DAC baseline and MPEG HE-AAC at equal or lower bitrates.

Word-aligned prosody is increasingly used to detect depression and posttraumatic stress disorder (PTSD), yet the stability and clinical relevance of these measurements remain unclear. We examine stability at two levels. At the word level, we ask whether a speaker produces the same word in a similar speaker-specific way across different utterance contexts. At the speaker level, we ask whether averaging prosody across many words produces consistent speaker scores across separate recordings. Drawing on principles of psychological measurement, we compare these representations in four corpora spanning acted, spontaneous, and clinical speech. Word-specific deviations show little cross-context recurrence (0.01-0.16), whereas speaker summaries are highly consistent, reaching 0.997 in ESD, 0.73-0.90 in MSP-Podcast, and 0.84-0.996 in DAIC-WOZ and PTSD-STOP. Despite this consistency, associations with depression or PTSD severity do not exceed 0.15; equivalence tests exclude effects of 0.30 in all analyses and 0.20 in DAIC-WOZ, while within-person associations with daily PTSD symptoms are near zero ($|\rho|\leq0.04$). Thus, aggregation yields stable speaker-level prosodic measures, but these measures do not detectably track depression or PTSD severity in the studied corpora. Code is available at https://github.com/jen900704/word-aligned-prosody.

Speech representation learning requires a discrete modeling interface while preserving high waveform fidelity. Existing approaches rarely achieve both: discrete tokens ease modeling but lose fine-grained acoustic information, while continuous latents preserve fidelity but require specialized generative objectives. We introduce SemiCodec, a novel Semi-discrete neural speech Codec with synchronized discrete tokens and continuous latents. Direct joint optimization may create a reconstruction shortcut through the continuous pathway, leaving the discrete tokens insufficiently grounded. To address this issue, we propose a Preserve-then-Ground training paradigm that first establishes a high-fidelity synthesis space and then grounds SSL-guided discrete tokens within it. Experiments show that SemiCodec outperforms fully discrete codecs in reconstruction quality and intelligibility while remaining competitive with continuous representations in acoustic fidelity and speaker preservation. Audio samples are available at https://pennyw1-se.github.io/SemiCodec-Demo-Page/

Recent alignment-free Text-to-Speech (TTS) models achieve high synthesis quality by removing explicit alignment modules and simplifying training. However, they often struggle in cross-lingual scenarios due to differences in character durations across languages, leading to duration mismatch and degraded speech quality. Existing approaches typically address this issue using explicit duration predictors, increasing model complexity. Systematic evaluation of TTS performance for Indian languages in cross-lingual settings remains under-explored due to the lack of standardized evaluation benchmarks. In this work, we construct a cross-lingual evaluation benchmark covering English and nine Indian languages and propose a language-aware inference-time duration estimation strategy for alignment-free TTS that incorporates language-specific duration-per character statistics without retraining or architectural modifications. We further develop heuristic formulations that refine duration estimation based on language-dependent duration variations. Experiments on bilingual and multilingual A-DMA models demonstrate improved speech naturalness, intelligibility, and speaker similarity while reducing duration mismatch.

State-of-the-art beat tracking is based on neural networks that model beats and downbeats as individual events, often followed by a separate Dynamic Bayesian Network (DBN) enforcing musical consistency. We propose a novel architecture that combines deep learning with differentiable theory-based modules: Specifically, we design a system that first predicts a tatum grid, then groups tatum events to beats and downbeats, explicitly modelling multiple options for the current tempo and time signature in a bar-pointer-like fashion. Our approach improves continuity metrics compared to existing DBN-free systems and is more versatile than the de-facto standard DBN, but still falls behind the best DBN-based methods on average. Given the many design choices and hyperparameters involved, we assume we have not yet explored the full potential of this idea and publish our code and invite others to explore this further.

User-generated audio collections frequently contain textual metadata such as titles, tags, and descriptions. Effectively leveraging these complementary fields requires accounting for both field het erogeneity and varying semantic relevance relative to the acoustic content. We propose FAM-Gate, a two-stage framework for metadata-enriched audio classification. In the first stage, field-aware metadata-view attention dynamically aggregates multiple metadata views using audio-guided weighting, explicit field-identity embeddings, a shared identity-initialized projection, and mean-anchored regularized aggregation. In the second stage, a prior-constrained vector gate performs bounded, dimension-wise fusion of audio and metadata representations centered around an acoustic prior. Ex periments on BSD10K, BSD35K, and FSD50K show consistent improvements in both hierarchical single-label and multi-label classification. Compared with matched random-view scalar-fusion base lines, FAM-Gate improves the hierarchical F-score by 2.53 and 5.72 percentage points on BSD10K and BSD35K, respectively, while increasing macro mAP and macro F1 by 5.07 and 10.25 percentage points on FSD50K.

Recent Speech Full-Duplex (FDX) models are capable of engaging in naturalistic conversation by listening and speaking at the same time. Building on this foundation, recent architectures introduce internal thinking streams to enhance their intelligence. To guarantee these streams process successfully in real time, existing methods enforce a fixed decoding rate, which assumes a hardware profile capable of a specific latency. We argue this static approach is severely suboptimal. Because internal thinking is intrinsically asynchronous, an ideal FDX model should sustain real-time conversational flow while allowing each stream to decode at its own hardware-adaptive rate. Thinking streams must process as fast as the hardware allows to deliver accurate answers faster. To solve this, we propose RUBATO, a novel multistream positional encoding that assigns independent clocks to entirely decouple the streams. This structural isolation empowers the model to zero-shot interpolate across any inference decoding speed. It drastically simplifies dataset construction by requiring training on only two extreme decoding speeds while smoothly adapting to any intermediate rate during deployment. RUBATO demonstrates consistent robustness across five spoken reasoning tasks. On Spoken GSM8K, a baseline model collapses to 1.7% accuracy at interpolated speeds. Conversely, RUBATO-Full sustains 48.7% to 51.1% accuracy across all rates, leveraging faster hardware to achieve a 6.5x shorter time-to-answer with a constant time-to-first-response. This advantage holds on BIG-Bench Audio and amplifies significantly during complex full-duplex interruptions. Specifically, when a user interjects while the model is actively speaking and reasoning, RUBATO-Full achieves 40.0% to 45.1% accuracy, whereas the baseline fails completely.

Recent advances in automatic pronunciation assessment (APA) have driven a shift from scripted to free-form assessment. However, existing methods often rely on multi-stage pipelines with forced alignment and separate feature extractors, limiting their practicality across response settings. In light of this, we propose OpenPA, a novel APA framework that leverages Whisper's cross-attention mechanism to jointly predict word- and utterance-level pronunciation scores without forced alignment. Through mixed-transcript adaptation and adaptive substitution weighting, OpenPA accommodates heterogeneous transcripts across closed- and open-response settings within a single model. Trained solely on speechocean762, OpenPA achieves the strongest overall performance among evaluated systems under in-domain closed and simulated open-response conditions, while maintaining its performance zero-shot on MultiPA. An additional cross-task evaluation on oral-cancer speech suggests its potential applicability beyond non-native speech.

Speaker verification (SV) systems now achieve very low error rates on standard public protocols, but those results do not fully describe deployment generalization. VoxCeleb-style evaluations remain useful anchors for English in-the-wild media speech. Scenario-specific evaluations and challenges are also valuable. However, neither setting directly shows whether the same system remains reliable across diverse deployment scenarios. We introduce OpenSVBench, a public benchmark for cross-scenario SV evaluation. Its fixed public core contains 26 trial sets organized into 11 scenario groups, with 9.15M verification trials, frozen target/non-target keys, and metadata for scenario-wise reporting. Around this core, the online leaderboard acts as a diagnostic evaluation interface rather than only a ranked list. It accepts score packages from new systems, evaluates them with frozen keys and common scoring, and reports trial-set breakdowns, scenario-group metrics, macro EER/minDCF, model-wise profiles, and scenario-wise comparisons. We evaluate 18 public checkpoints spanning x-vector, ECAPA-TDNN, CAMPPlus, ResNet and ERes variants, ReDimNet2, SAM-ResNet, and pretrained Transformer speaker recipes. The results show clear cross-scenario gaps: systems with similar macro EER can have different scenario profiles, strong performance in one scenario does not guarantee strong performance in another, and same-scenario trial sets can still expose source and protocol gaps.

Given the challenges associated with obtaining large volumes of real-world audio, the advancement of Automatic Speech Recognition (ASR) systems increasingly depends on synthetic data. This study focuses on the call center-banking domain, providing a targeted analysis of ASR models in this context. We evaluate real and synthetic datasets, using Word Error Rate (WER) and Character Error Rate (CER) as metrics, and employ a speech quality model to assess the impact on ASR performance. Our research compares Text-to-Speech (TTS) and advanced voice conversion methods, including KNNVC, Seed-VC, and Vec2Wav2, revealing significant improvements in speech quality and ASR accuracy, particularly with Seed-VC. Additionally, our domain-specific experimentation provides insight into the unique challenges and opportunities that arise when applying ASR technologies to industry-relevant settings. This highlights the potential of voice conversion technologies to enhance ASR systems, guiding future research in diverse linguistic scenarios, and paving the way for the broader application of ASR innovations across various fields.

Vocal intensity conversion renders a recorded utterance at a different vocal intensity, rather than simply scaling its amplitude, while preserving its content and speaker identity. Existing systems control intensity through a reference utterance or a few discrete categories. Instead, we control the conversion with any target sound pressure level (SPL) on a continuous dB scale. A lightweight converter, trained as a conditional cycle-consistent GAN without parallel data, transforms the features of a frozen speech encoder, from which a pretrained vocoder resynthesizes the speech. We compare three representations, log-mel spectrograms, WavLM, and SpeechTokenizer, and introduce an objective protocol that measures the achieved intensity change against the target one on a calibrated corpus. With WavLM, the converter achieves a measured 83% of the target change in intensity and also shifts the fundamental frequency. A perceptual evaluation confirms that perceived vocal intensity increases with the target level on amplitude-normalized stimuli.

Large Audio Language Models (LALMs) achieve strong performance on speech-text tasks, but standard audio encoders produce hundreds of tokens for even short audio clips. Because self-attention scales quadratically in sequence length, this inflates inference-time compute substantially. Current training-free token compression methods prune or merge tokens based solely on coarse attention or feature similarity. However, treating audio tokens as generic, isotropic features smooths over spectral transitions, destroying fine-grained phonetic transitions essential for decoding speech. In contrast, speech signals exhibit highly structured temporal patterns where information density peaks at acoustic transitions. We propose AudioTETRIS, a training-free, two-stage framework that uses the audio encoder's representation geometry as an acoustic prior for token importance before applying attention-based compression. Instead of discarding encoder structure after the encoding stage, AudioTETRIS propagates this prior directly into the LLM decoder to compute a per-token fused score. This score drives progressive top-$k$ token selection across decoder layers without requiring additional training or auxiliary models. Evaluated on Qwen2-Audio-7B, Phi-4-Multimodal, and Audio Flamingo 3, AudioTETRIS retains near-uncompressed performance on several speech tasks using only 25% of audio tokens, matches or surpasses full-token accuracy at moderate compression rates, and consistently outperforms existing training-free baselines.

Autoregressive Transformer and Conformer models have achieved state-of-the-art performance in automatic speech recognition (ASR). However, their sequential token generation mechanism imposes a significant bottleneck during decoding, resulting in substantial inference latency. While speculative decoding has emerged as a promising lossless acceleration technique, existing ASR-oriented approaches either suffer from low draft acceptance rates due to suboptimal draft quality or rely on specially designed external draft models, which limit their generality and practicality. To overcome these limitations, we propose EAGLE-ASR, a novel framework that accelerates ASR inference via EAGLE-style speculative decoding while strictly preserving lossless output quality. Specifically, we introduce Speech-Draft, a lightweight autoregressive module that serves as a universal draft model by jointly leveraging encoder acoustic features and hierarchical decoder representations from the target ASR model. Furthermore, we adopt a multi-step generation strategy with self-distilled data to train the Speech-Draft, enabling it to better approximate the target model's output distribution and maintain causal consistency during multi-token generation. Extensive experiments on representative Whisper and Conformer AED models demonstrate that EAGLE-ASR achieves significant lossless speedups of up to $1.9\times$ and $2.4\times$, respectively, consistently outperforming existing speculative decoding baselines for ASR. Our proposed framework further generalizes effectively to multilingual ASR and speech translation tasks, establishing EAGLE-ASR as a practical and universally applicable solution for accelerating modern ASR systems without compromising recognition accuracy.

Target speaker extraction (TSE) is a key frontend for enabling audio large language models (LLMs) to recognize and understand a designated speaker in overlapping speech. However, multilingual mixtures exhibit heterogeneous extraction demands, causing fixed-depth systems to either under-process challenging inputs or waste computation on inputs that have already converged. Existing TSE frontends generally apply uniform computation, while Audio-LLM adaptation methods neither explicitly model language-conditioned extraction demands nor preserve plug-and-play compatibility with frozen downstream models. We therefore propose LATTE, a lightweight 5.2M-parameter frontend that combines an acoustic Difficulty Estimator, a task-adaptive phonology-informed prior learned from multi-depth TSE gains, and an Adaptive Hourglass Extractor to route each mixture without requiring language identity at inference. On PolyExtract, LATTE achieves a macro SI-SNRi of 15.07 dB at 0.30 GFLOPs, retaining performance within 0.09 dB of a 0.53-GFLOP fixed-depth model while reducing computation by 43.4%, and lowers the average WER from 16.8% to 14.6% relative to TF-GridNet. By jointly modeling sample-level extraction demands and phonological structure, LATTE provides a practical framework for quality-aware, compute-efficient multilingual TSE and offers a scalable path toward robust Audio-LLM deployment.

Speech recognition on low-cost microcontrollers is, in practice, limited to keyword spotting or a fixed list of commands. We present Oído, an inference engine that runs open-vocabulary English speech recognition entirely on an ESP32-S3 (240 MHz dual-core Xtensa LX7, 512 KB of SRAM, 8 MB of PSRAM, 16 MB of flash, no neural accelerator). Oído executes a 13M-parameter Conformer-CTC model with int8 weights, per-frame dynamic int8 activations and int8 attention on the chip's SIMD instructions, with a schedule designed around flash and PSRAM bandwidth. On LibriSpeech test-clean/test-other the on-device arithmetic reaches 3.70%/8.23% word error rate (WER), within 0.12 points of full precision, and 3.33%/7.15% with an on-chip 1.3M-parameter language model; a transducer head gives 2.99%/6.67% and an int4 variant fits in 8.3 MB. This is 2.3-2.8x fewer errors than the vendor's recognizer on the same chip, and fewer than Whisper tiny.en and Moonshine tiny in full precision on a laptop, including in real noise and reverberation. Measured on a physical board, the firmware runs at 1.97x the duration of the audio on one core, with transcripts identical to an instruction-exact emulator: accurate, but not yet real time (a 4 s command is transcribed about 8.7 s after it ends). The two-core schedule gives wrong results on silicon and is disabled. A streaming variant shows partial text while the speaker talks and delivers the final text 4.1-11.1 s after clips of 3.6-10.6 s (7.2-22.4 s in utterance mode) at 4.86%/11.03% WER; a Spanish model makes 34-60% fewer errors than multilingual Whisper tiny. Code and models are released.

Recent accent classification systems can achieve high accuracy in standard in-corpus evaluations. However, when speakers overlap between the training and test sets, such results may overestimate model robustness compared with real-world deployment scenarios. This study revisits the robustness of DistilHuBERT-based accent classification by using L2-ARCTIC and multiple external English accent datasets. The comparison covers file-level and speaker-disjoint splits, along with balanced external-data training and teacher-student pseudo-labeling. The large performance gap between in-corpus and external evaluations suggests that file-level evaluation substantially overestimates performance, while speaker-disjoint evaluation can provide a more realistic estimate of the generalization to unseen speakers. In this study, models trained on external data alone do not reliably generalize to L2-ARCTIC. Among the tested configurations, balanced external-data training offers the most promising supervised setting. The pseudo-labeling experiments further indicate that confidence-based filtering alone is insufficient under domain mismatch. Overall, this study suggests that robust accent classification evaluation should simultaneously include speaker-disjoint evaluation and cross-corpus testing.

Earnings call transcription presents a realistic continual learning scenario where Automatic Speech Recognition (ASR) systems are updated with new recordings but risk catastrophic forgetting of previously learned financial speech patterns. Adapting Whisper-small from a general-domain speech corpus to the Earnings22 benchmark reveals the opposite: all evaluated methods exhibit improvement on held-out financial speech from the same domain during adaptation rather than forgetting. This task complementarity makes regularization counterproductive. Nevertheless, when regularization is warranted, a layer-wise mini-Kronecker-Factored Approximate Curvature (mini-KFAC) approximation seems to achieve lower word error rates (WERs) and more stable optimization than standard diagonal Elastic Weight Consolidation (EWC). The results suggest that task complementarity should be assessed before applying continual learning constraints, and prefer mini-KFAC EWC over diagonal EWC when needed.

Multimodal emotion preference learning uses pairwise comparisons to identify which open-ended emotion descriptions align more closely with human judgments. However, existing methods primarily focus on the final preference decision and provide little explicit supervision for localizing and correcting erroneous content. To address this limitation, we propose EmoGavel, a Grounding-Aware reinforcement learning framework with Verifiable rewards for multimodal Emotion preference Learning. We first introduce grounded supervised fine-tuning (Grounded SFT), which exploits controlled-error construction information to jointly learn preference prediction, error localization, repair, and missing-evidence recovery. We then optimize these verifiable outcomes with grounding-aware group relative policy optimization (GA-GRPO), combining preference-gated grounding rewards with dynamic zero-variance rescue to prioritize preference correctness and recover learning signals from all-wrong groups. Experiments on EmoPrefer and controlled-error evaluation sets show that EmoGavel improves both natural preference prediction and controlled-error discrimination while maintaining strong error-grounding performance. Our code and model checkpoints will be released once the paper is accepted.

Speech anonymization is evaluated with automatic speaker verification attackers whose performance depends on the attribute distribution of both their training and evaluation data, which may bias privacy estimates for some speaker groups. To alleviate this issue, we introduce VoxAnon, a dataset built from VoxCeleb1 and 2 with training and evaluation splits balanced across gender, age, accent, and pitch. We extract and cross-check attributes using two complementary methods, balance trial lists per subgroup, and draw test/enrollment utterances from distinct video sources in order to limit linguistic and session bias. Comparing VoxAnon with unbalanced training and test data across four attackers and three anonymizers, we show that attribute imbalance changes which subgroups appear hardest to re-identify, in directions that depend on the attribute. In the balanced setting, residual disparities persist, mostly tied to accent and pitch, and depend on the anonymizer/attacker pair. Code and trial lists are publicly available.

This paper describes our systems for the Track 1 time-stamped speaker-attributed ASR task of the SmartGlasses Challenge. To handle overlapping speech, background noise and far-field conditions in two-person conversations, we employ SoulX-Transcriber, a multimodal large language model that takes full-length audio as input and generates structured transcriptions. In addition, we investigate the effect of speech frontend processing, including a rule-based Guided Source Separation method and a data-driven speech separation model, both of which improve performance on the validation set. Considering inference stability, our final submission uses the fine-tuned SoulX-Transcriber system, achieving a tcpCER of 6.22 on the test set.

Open audio-language models are used as judges of generated music: given a clip and its text prompt, the model rates the clip, and the judge is validated by how well its ratings correlate with expert ratings. The correlation does not show what the judge responds to: the music, the text prompt, or the way the question is asked. We propose four probes that test each of these, three of them read against a control computed from the human ratings: removing the audio, replacing the clip with another model's clip for the same prompt, swapping the order of two clips in pairwise judging, and rewording the instruction. We apply them to five open judges on all 2,748 clips of MusicEval, and the last two to the SongEval songs. Without audio, no judge exceeds what the prompt alone predicts from the human ratings, and for prompt alignment every judge falls below it. With audio, every judge follows the clip it hears when rating prompt alignment, two do when rating musical quality, and the strongest judge's quality score tracks the experts' alignment rating more than their quality rating. The pairwise verdicts of the four smaller judges follow the position of a clip in the pair (its slot) rather than its quality on both rating sets. For musical quality, rewording the instruction moves the correlation of most judges as much as changing the judge. MusicEval's prompt assignment bounds any prompt-only judge at the model level. We recommend reporting the four probes with every music judge.

We study query-conditioned multi-instance open-vocabulary sound event localization (QC-MI-OVSEL), where a system must recover all active physical sources matching a natural-language category query, including when the queried category is absent. We introduce the QC-MI-OVSEL benchmark, a 40,000-scene first-order Ambisonics benchmark with 80,000 balanced present/absent query views over 83 classes, including 17 novel classes excluded from training and development. Each frame is annotated with the activity and 3-D direction of every query-matching source, with up to three simultaneous targets. We further propose Q-RSet, which uses query-conditioned source slots and responsibility-based evidence aggregation for source-set prediction. Experiments show improved source recovery, localization recall, cardinality estimation, exact-set recovery, and absent-query suppression over complementary baselines. Analysis indicates that novel categories, paraphrased queries, and higher source multiplicity mainly impair source discovery and set completion, while directional accuracy remains comparatively stable for successfully recovered sources.

Recent advances in speech synthesis technology have enabled the generation of speech with diverse characteristics. One such characteristic is a foreign accent. In our preliminary experiments, we found that UTMOS, a widely used automatic speech quality assessment system, tended to assign lower scores to speech with stronger foreign accents, even though it is produced by a human speaker. This observation raised a question: How do humans evaluate the naturalness of foreign-accented speech? We conducted a large-scale subjective evaluation involving both native and non-native listeners under different evaluation instructions. The results showed that humans also considered accentedness when evaluating naturalness, but that defining naturalness in terms of human-likeness and selecting raters who have greater exposure to the target accent may enable more appropriate evaluation. This study aims to provide a starting point for future discussions on fairer and more appropriate evaluation of synthetic speech with a foreign accent.

Large audio-language models (LALMs) frequently exhibit object hallucination, affirming absent sounds. The dominant mitigation, Audio-Aware Decoding (AAD), contrasts logits under real audio against an all-zero "blank" waveform, flagged by prior review as out-of-distribution. We introduce a black-box causal test that isolates the claimed evidence itself: localize a claim's evidence region via CLAP similarity, remove it by spectral gating, and compare against a random-window control across a 2x2x2 grid on three LALMs. On Qwen2-Audio-7B, localized ablation's raw AUROC trails blank's (0.613 vs. 0.721); this is a duration confound, not a localization failure, since blank removes the entire clip versus the grid's 2-4 s. Holding duration fixed, localization beats random removal at every cell on both models. Raw detection value remains model-dependent, especially on Audio Flamingo 3. The same evidence corrects decoding: contrastive decoding combining blank and language-prior contrasts, gated and ensembled, raises Qwen2-Audio accuracy from 57.1% to 71.9%, beating AAD alone. A unified causal-grounding profile does not transfer between models, indicating different grounding regimes. The source code and a detailed report of our findings, including identified limitations, are available at https://github.com/khalooei/causa.