arXane › Computer Science › Computer Science
Wednesday, 4 November 2026: 9 new papers, 4 cross-lists and 7 replacements in Sound (cs.SD), in arXiv's order, with full abstracts. The same papers as arXiv's own listing, set as a newspaper; every title links to arXiv.
Respiratory sound classification from auscultation recordings is confounded by heterogeneous acquisition conditions, including recording device, auscultation site, and patient characteristics, that introduce systematic variability unrelated to the underlying pathology. Although such metadata is often available, existing approaches either ignore it or rely on na\"ive fusion strategies that do not model how distinct metadata sources influence different components of the acoustic representation. We propose a family of metadata-conditioning mechanisms for the Audio Spectrogram Transformer (AST), progressing from gated residual fusion to Feature-wise Linear Modulation (FiLM) adapted to the Transformer architecture. We further introduce two novel variants: Token-Aware FiLM (TAFiLM), which applies distinct modulation to classification and patch tokens, and Soft-Factorized FiLM (SoftFiLM), which learns soft masks to align metadata factors with disentangled feature subspaces. Experiments on the ICBHI benchmark demonstrate consistent gains across all conditioning strategies. SoftFiLM achieves 64.11% on the official 4-class task (+3.94% over the baseline) and 72.40% on the 2-class task, establishing a new state-of-the-art on the 2-class task and competitive performance on the 4-class benchmark.
Multimodal language models continue to improve on broad music understanding benchmarks, but which musical features can they perceive and reason about? We investigate this question by introducing PianoBench, a benchmark that tasks language models with reasoning about musical performances across congruent acoustic and symbolic input representations. PianoBench is derived from our new multimodal, training-scale dataset comprising more than 20,000 hours of interleaved transcripts of music and spoken musical commentary from instructional piano videos. Open-weight audio-language models achieve at most 52.8% accuracy on PianoBench, and at best only modestly outperform their accuracy on a no-audio control. In contrast, frontier models using chain of thought achieve up to 93.2% accuracy when provided transcripts of music as text, exposing a gap between perceiving musical features from audio and reasoning about musical events when they are given explicitly. We then test whether mid-training on interleaved musical instruction can improve question answering without chain of thought, raising Qwen 3.5 9B Base's zero-shot accuracy on PianoBench with music transcripts from 42.8% to 56.0% after instruction tuning. These experiments suggest that instructional videos may represent an overlooked source of multimodal supervision for music. Available: https://gclef-cmu.org/pianobench.
Competing speech can cause excess emissions in a connectionist temporal classification (CTC) recognizer. We penalize the amount by which the noisy student's expected collapsed token count exceeds a frozen teacher's count on the paired clean utterance. The one-sided objective caches one scalar per utterance and adds no inference module. With 147,456 trainable parameters in wav2vec 2.0, it improves competing-speech word error rate (WER) over matched CTC-only adaptation across three development seeds. On a locked LibriSpeech test, speech-interference WER falls from 45.198% to 44.323%, while clean WER changes by 0.004 percentage points. Unseen VCTK interferers yield a 0.307-point improvement. Ablations favor collapsed-count control over frame-mass regularization. Longer training increases the speech gain but worsens noise and music performance, identifying an insertion-deletion tradeoff rather than uniform robustness.
Spatial audio question answering requires associating sound identities with locations and reasoning about their relations. This paper introduces SciPhi-Chat, a spatial audio agent that answers questions, conducts multi-turn dialogue, and calls tools on first-order ambisonic recordings. The accompanying MASU benchmark contains 12,000 samples across 42 task categories, the widest task coverage of any spatial model, with real recordings and simulated audio scenes. SciPhi-Chat achieves a score of 44.9%, 7.4 percentage points above the previous state-of-the-art. On the external ASA-Bench, SciPhi-Chat achieves 74.5% azimuth and 60.0% elevation accuracy, the highest among locally evaluated models. Removing reasoning supervision reduces ASA-Bench accuracy by 17.4 percentage points. All materials will be released via https://github.com/microsoft/sciphi-chat
Modern text-to-speech (TTS) systems can synthesize natural and expressive speech, yet a practical speech foundation model must go beyond naturalness to offer stable quality, precise controllability, and a broad set of composable capabilities within a unified framework. We present TTSonic, a unified speech generation model that supports zero-shot TTS, voice cloning, prosody cloning, instruction following, voice conversion, and cloning from noisy reference speech. Rather than relying on separate task-specific models, TTSonic accommodates diverse conditioning signals, including text, natural language instructions, reference speech, and text-described timbre, within the same model framework. Specifically, we develop a semantic--acoustic disentangled speech tokenizer through ASR-supervised semantic learning and reconstruction-guided generative refinement. Serving as the representational foundation of our unified framework, the tokenizer enables the Text-to-Semantic autoregressive model to focus on semantic and prosodic planning, while the acoustic flow-matching model specializes in high-quality speech rendering under diverse acoustic and timbre conditions. Building on this framework, we incorporate CoT-guided instruction-following planning and Text2Timbre voice design to support controllable speech generation. Experiments across zero-shot synthesis, multilingual and cross-lingual TTS, instruction following, voice design, voice conversion, and noisy-reference cloning demonstrate the competitive performance and broad applicability of TTSonic.
Personalized neural audio watermarking systems embed user-specific payloads into audio to trace the user associated with a leaked or misused copy. Existing evaluations mainly assume access to a single personalized copy, but real-world distribution may expose multiple copies of the same recording. We find that simple sample-wise averaging breaks traceability: averaging only two copies produces tracing-failure rates of 87.7-99.7% across the five systems while largely preserving audio quality. Decoder outputs show distinct responses to coalition payload composition across systems, with high majority agreement in TimbreWM but substantially lower agreement in WMCodec. In a more sophisticated payload-aware setting, mixture weighting can further steer the decoder toward a selected nonmember, achieving a target success rate of 92.3% in TimbreWM. Preliminary confidence screening rejects many collusion-induced assignments while retaining most valid single-copy outputs. Code and results are available at https://github.com/Silwan5z/neural-audio-watermark-collusion-attack. This work highlights multi-copy traceability as an additional evaluation dimension for personalized neural audio watermarking alongside conventional single-copy robustness.
In this paper, we propose a novel Text-to-Speech (TTS) framework designed to perform adversarial attacks against speech deepfake detection systems. Given a TTS model and a detector capable of classifying its outputs as synthetic, our approach fine-tunes the vocoder by integrating the detector into the TTS training loop. Unlike traditional adversarial methods that rely on post-hoc perturbations or external noise injection, our approach enables the generation of synthetic speech that is intrinsically adversarial, capable of lowering the target detector's reliability without requiring additional post-processing steps. We evaluate the framework in both white-box and black-box settings, showing its effectiveness against known and unseen detectors. We also propose a multi-detector strategy that allows the system to simultaneously learn to deceive multiple detectors, showcasing the scalability of the method. Experimental results confirm the efficacy of our approach, which has the potential to impact forensic security and detection countermeasures. At the same time, subjective listening tests reveal that the proposed method also enhances synthesized speech quality.
Large Audio Language Models (LALMs) enable generative reasoning over complex acoustic scenes but often exhibit prediction instability. We show that this variability can serve as a reliability metric via inference-time self-consistency and confidence-aware aggregation. Using a log-probability--based scoring framework, candidate answers are weighted by model likelihoods, leveraging reasoning diversity and self-certainty without additional training. Experiments on four LALMs, including Qwen3-Omni and Audio Flamingo 3, demonstrate that confidence-aware aggregation improves accuracy and calibration, with additive methods like Borda voting outperforming classical Majority Vote, especially under high disagreement. This approach transforms LALM variability into a source of reliability for scalable and robust multimodal audio understanding. Project code is available at https://github.com/michelolzam/lalm-self-consistency.
Small multirotor drones produce a characteristic acoustic signature, and microphones offer a low-cost way to detect them. High accuracies have been reported on public drone audio benchmarks, often with random clip-level splits of datasets built from a few long recordings and pieces of existing sound collections, so pieces of one source can appear in both training and test data. We study the binary task of the public DroneAudioDataset (11,704 one-second clips). Its 1,332 drone clips come from 257 five-second chunks of 7 recordings, and 96.4% of its non-drone clips are pieces of the 2,000 clips of the ESC-50 sound dataset. A grouped, source-disjoint protocol (five folds, three repeats) is compared with a random split. Each clip is described by 57 hand-crafted acoustic descriptors, reduced by a hybrid minimum-redundancy maximum-relevance and ReliefF selection or by NSGA-II multi-objective search, and by a 1024-dimensional embedding from the pretrained YAMNet network. Random Forest and XGBoost learners on both views form a stacking ensemble. Under the grouped protocol, the two-view stack reached a mean accuracy of 95.29% (SD 1.62 points over 15 test folds), a specificity of 98.29% and an area under the ROC curve of 0.990, and it correctly classified 98.78% of unseen ESC-50 source clips as non-drone. The random split overstated accuracy by 1.89 to 3.22 points for the four models. Adding YAMNet embeddings to the hand-crafted view raised accuracy by 1.11 points (higher in 12 of 15 folds). Hybrid selection matched the accuracy of all 44 baseline descriptors with 16 to 18 features, and NSGA-II retained 4 to 6 features at lower accuracy. The stack took 51.0 ms per clip on one CPU thread. Source-disjoint evaluation is needed for a fair benchmark on this dataset, and fusing hand-crafted descriptors with pretrained embeddings gives a small but consistent gain at modest computational cost.
Cardiac auscultation traditionally detects the presence of a heart murmur, but clinicians further characterize it by timing, shape, pitch, quality, and grade. Most automatic phonocardiogram analysis systems stop at binary detection and do not estimate this richer description. In this work, we formulate automatic murmur characterization as five patient-level classification tasks and compare three multi-spot strategies on recordings from the aortic, pulmonary, mitral, and tricuspid auscultation locations. First, classical and deep-learning classifiers are trained independently at each location, and predictions are combined via a maximum-severity late-fusion rule. Second, a synchronous multi-input CNN jointly processes one representative segment per location for a single task. Third, a unified multi-input, multi-output CNN predicts all five attributes from one shared multi-spot representation. Using 735 patients from the CirCor DigiScope dataset with complete four-spot recordings, the best late-fusion configurations reach an unweighted mean sensitivity between 85.56% and 93.81% across the five tasks, with class-specific recalls between 80.00% and 100.00% for every relevant category. The multi-input CNN trails late fusion by 7-16 points depending on the task, while the multi-output model reaches values between 72.59% and 79.40%, estimating all five attributes from one shared representation. We further show that naively balancing training data by adding all minority-class segments degrades patient-level performance under the maximum-severity rule, since it introduces non-representative, noisy segments not filtered by the clustering-based segment selection used for evaluation. These results indicate multi-spot murmur characterization is feasible with both fusion-based and jointly learned architectures, outlining a path toward automatic, structured murmur reports complementing binary screening.
We present HyVox, a hybrid neuromorphic--von Neumann system for low-energy open-set speaker identification under tight energy budgets. HyVox splits the workload at the speaker embedding. A neuromorphic processor embeds the speech with sparse, event-driven computation, and a von Neumann microcontroller holds the speaker gallery and makes every decision: it scores queries by cosine similarity, rejects unknown speakers, and enrolls new users by writing a prototype, with no weight update. On the neuromorphic processor, a hierarchical temporal abstraction (HTA) SNN backbone learns a synaptic time constant per leaky integrate-and-fire (LIF) neuron, allocating 832 neurons across fast, medium, and slow timescales. A 32-neuron spiking embedding extractor (SEE), pretrained on episodes that each enroll new speakers and reject outsiders, turns the backbone output into an embedding for speakers unseen in training; its 32 spike counts are the only interface between the two processors. To fit the chip's synapse and integer budgets, we compress the backbone 6.5x to 29.0k synapses, train it for integer bit-shift dynamics, and fine-tune it with chip-accurate spiking training (CAST). CAST trains through a differentiable replica of the chip's integer neuron update, derived from its cycle-accurate simulator, so that gradient descent optimizes the exact weights and thresholds the chip runs; it doubles the top-1 accuracy of the mapped backbone on unseen speakers. On 20 Common Voice households of unseen speakers, HyVox reaches 99.6\% top-1 and 5.5\% rejection EER with 2 members and 94.9\% and 13.2\% with 20, and the network deployed on the neuromorphic chip matches its trained model. Each inference on a 3-s window takes 0.96 s and 1.63 mJ of dynamic energy, 42\% lower latency than the fastest embedded ANN baseline and 13.5x less dynamic and 5.8x less total energy than the most energy-efficient one.
Audio language models are meant to answer questions from the audio itself, but an hour-long recording rarely fits in their context, and the answer often depends on a few seconds of it. Transcribing the recording first makes it fit but discards sounds and voices, while systems that return to the audio must be trained for this behavior. We propose ReListen, a training-free method in which a frozen audio language model finds its own evidence before answering. In a Jev-style scan, the model scores each 20-second window by the probability that it helps answer the question, and it then answers from the highest-scoring minute. This minute overlaps the annotated evidence far more often than transcript retrieval, for 88.5% of VoiceGiraffe questions. Without any change, the same scan also grounds events on LAT-Bench as well as a model trained for temporal grounding. Answering from it matches or exceeds hearing the whole recording with a small fraction of the audio, and on VoiceGiraffe it exceeds a transcript of the whole recording by 9.2 points. The selected evidence is also worth hearing rather than reading. Given the same excerpts, reading their transcript matches listening on speech content, but listening is clearly more accurate on questions about sounds and voices. Moreover, ReListen answers faster than every transcript-based system, and its memory stays constant even for recordings too long to hear whole on one GPU. Therefore, a frozen audio language model can search hour-long recordings by itself, without a transcript or additional training.
A speech deepfake detector estimates whether a recording is synthetic, but a forensic analyst also asks which system produced it, whether recordings share a source, and what evidence supports each conclusion. We present VeriSpoof, an audio language model that unifies five tasks with shared weights: deepfake detection, generator attribution (generator, acoustic model and vocoder), recording-condition analysis, synthesized-span localisation, and source verification. Each decision is supported by grounded rationales whose statements can be independently verified. Answers are scored by sentence likelihood, allowing detection and source verification to be thresholded without separate heads. Rationale statements can be checked against waveform measurements, span labels, or generator metadata. Counterfactual examples train decisions to follow the stated evidence. Across the 14 Speech-DF-Arena test sets, VeriSpoof-Plus achieves 2.14% pooled EER, ranking third of 28 systems and first among open-weight systems. VeriSpoof-Base, trained only on official training splits of eight corpora, achieves 4.43% EER, less than half that of the next open-weight detector. With the same weights, VeriSpoof-Plus verifies whether recordings from generators unseen in training share a source at 6.2% EER. Controlled studies show that the audio representation governs detection, the LLM governs recording comparison, and source supervision is necessary for neural-codec fakes. Statement-checkable rationales catch partial fakes missed by one-step answers (99.6% vs. 78.4% on PartialEdit) and determine decisions under edits unseen in training (99.5%). Training updates only 29.7M of 2.85B parameters, fitting a single 24 GB GPU with gradient checkpointing. We release both models, inference code, a demo, and VeriSpoof-Bench, the evaluation suite covering these questions.