Papers

Audio-visual instance segmentation (AVIS) requires accurately identifying and tracking individual sounding objects with pixel-level masks. Existing methods struggle to match overlapping acoustic events with visual instances and handle asynchronous audio-visual dynamics. Therefore, two critical questions arise: how can a model establish precise correspondence between overlapping sound sources and visual instances, and how can a model maintain robust tracking when audio and visual signals are temporally misaligned?This paper proposes Hear to See (H2S), addressing these challenges through two mechanisms. The Acoustic-Semantic Projector (ASP) disentangles mixed audio and establishes hierarchical correspondence from semantic to spatial domains. The Asynchronous Dynamics Modulator (ADM) adaptively adjusts state transitions via audio-modulated Mamba, prioritizing current information during dynamic variations and maintaining continuity in stable periods.Experiments on AVISeg show H2S achieves SOTA performance, attaining 48.54 mAP with a COCO pretrained ResNet50 and surpassing the previous by 7.8%. The code will be open-sourced once the paper is accepted. The source code will be publicly available at https://github.com/leiyeliu/H2S.

Reference-conditioned melody-preserving lyric editing replaces words while retaining a performance's timing, singer identity, and naturalness. Continuous-latent autoregression avoids finite codebooks and offers stepwise generation with learned stopping. Editing creates a conflict absent from ordinary reconstruction: training pairs reference cues with original lyrics, whereas inference asks revised lyrics to override source-lyric-correlated cues; one source-following patch can propagate through AR history. We introduce CLASVS. Its State-Control-Transition (SCT) routing keeps target-lyric and reference-melody controls persistent, returns semantic feedback on phonetic progress to the causal planner, and confines the previous latent patch to the local Transition. Progressive State-Control Grounding (PSCG) learns this contract through paired-edit-free, content-consistent Mandarin reconstruction. On two Mandarin benchmarks, CLASVS improves all four operations over discrete-AR Vevo2 and reduces macro-PER by 46.2%, while maintaining melody, singer similarity, and perceptual quality. Together, these results establish a strong continuous-AR operating point for score-annotation-free lyric edits and a basis for broader stepwise control. Audio demonstrations are available on our project page: https://piedpiperg.github.io/clasvs-demo/.

Spoken emotional dialogue requires a model to understand a user's spoken input and generate a response that is both semantically appropriate and emotionally expressive. This is challenging because communicative intent may be stated explicitly in lexical content or conveyed more implicitly through paralinguistic cues, which can complement or diverge from the words themselves. However, two limitations constrain progress in this area: the scarcity of benchmarks that distinguish these intent expressions, and the lack of reinforcement learning objectives that jointly account for response quality and emotional expression. To address the lack of suitable benchmarks, we introduce ParaIntent, a Chinese benchmark comprising 14 intent categories with balanced explicit and implicit samples, together with a multidimensional evaluation protocol covering intent fulfillment, response quality, and emotional expression. For policy optimization, existing approaches either use a shared objective for text and speech or apply reinforcement learning to only one modality, leaving modality-specific learning signals entangled within policy optimization. Motivated by this, we propose Acoustic-Lexical Decoupled Policy Optimization (ALPO), which computes independent textual and acoustic advantages and routes them to the corresponding text and speech tokens within a unified rollout. Under identical reward functions and training budgets, ALPO improves over standard GRPO on most automatic metrics and achieves the best subjective results among the fine-tuned variants, with particularly clear gains in emotional expressiveness on both the synthetic and human-recorded test sets.

What is music style? Though often described using text labels such as "swing," "classical," or "emotional," the real style remains implicit and hidden in concrete music examples. In this paper, we introduce a cross-modal framework that learns implicit music styles from raw audio and applies them to symbolic music generation. Inspired by BLIP-2, our model leverages a Querying Transformer (Q-Former) to extract style representations from a large, pre-trained audio language model (LM), and further applies them to condition a symbolic LM for generating piano arrangements. We adopt a two-stage training strategy: contrastive learning to align auditory style with symbolic expression, followed by generative modeling for music arrangement. Our model generates piano performances jointly conditioned on a lead sheet (content) and a reference audio example (style), enabling controllable and stylistically faithful arrangement. Experiments demonstrate the effectiveness of our approach in piano cover generation, style transfer, and audio-to-MIDI retrieval, achieving substantial improvements in style-aware alignment and music quality.

While music generative models have recently gained significant attention, how they can be effectively integrated into live music performances still requires further exploration. This paper presents Calliphony, a calligraphy-driven interface for real-time generative music performance. Specifically, we build a low-latency pipeline that captures brush motion with an attachable sensor and maps it to control signals for real-time symbolic music generation. Using a generative model, the system produces multi-track MIDI in performance settings, while brush-derived control signals constrain event timing and activate additional musical layers. The generated melody is then extended with real-time harmony and additional voices, and finally rendered through a DAW for live staging. Calliphony contributes: (1) a performance-oriented prototype that uses calligraphic motion as an external control layer for a real-time symbolic music generation model, controlling note density, pitch constraints, and accompaniment-layer activation; and (2) a cross-modal performance scenario that extends calligraphy beyond a primarily visual practice into an audiovisual, AI-assisted setting.

Synthesizer inversion is challenging for two main reasons: 1) Distinct parameter configurations can produce perceptually similar sounds. 2) Parameter-space losses often fail to reflect rendered audio similarity, while the synthesizer being a non-differentiable black box prevents simple audio-domain supervision. To address the one-to-many mapping induced by the first challenge, we formulate synthesizer inversion as conditional generation over discrete synthesizer parameters and use masked discrete diffusion as the generator. This treatment additionally avoids the fixed-order assumption of autoregressive models and the continuous-relaxation mismatch of flow matching when modeling categorical synthesizer controls. To address the second challenge, we further fine-tune the model with GRPO-style audio-domain rewards computed from rendered outputs. Experiments on Dexed show that, after supervised training, the discrete diffusion model is competitive with autoregressive and flow-matching baselines, and reward-based fine-tuning further improves out-of-domain audio matching performance. Code and demos are available at: https://github.com/DDSynth-RL/DDSynthRL.

Neural audio codecs serve as fundamental tokenizers for LLM-based audio generation. While semantic priors are widely exploited to enhance linguistic intelligibility, the integration of explicit acoustic priors remains underexplored, limiting synthesis fidelity in frequency-sensitive domains. To address this gap, we introduce MeloCodec, a novel framework designed to effectively incorporate melodic priors, a critical form of acoustic information for singing. To address the optimization instability typically caused by the direct fusion of such explicit priors, we propose a Tokenize-then-Fuse paradigm that pre-trains a discrete melodic branch to lock in structures before feature fusion. To robustly realize this paradigm, we further propose a two-stage training strategy that prevents codebook collapse and ensures stable convergence. Experiments show that MeloCodec outperforms baselines in singing voice representation, improving pitch consistency and enabling controllable pitch manipulation with minimal timbre degradation.

This technical report describes Botnoi Group's methodology and results for rapidly fine-tuning the open-source NVIDIA Canary 180M Flash and NVIDIA Canary 1B Flash multitask models for speech-to-text tasks using the NVIDIA NeMo framework, with a focus on telephony-grade audio. To support this adaptation, we construct a telephony-oriented fine-tuning dataset from live voicebot system recordings and prompted speech with telephony-oriented augmentation. We evaluate four targeted experiments-language adaptation (Thai), telephony robustness, domain-specific jargon (names and addresses), and latency-using character error rate (CER) for accuracy and real-time factor (RTFx) for inference speed. Results show that fine-tuning substantially improves recognition in noisy telephony environments, reducing CER from 23.31% to 9.04% on BOTNOI telephony data, and further improves business-critical names and addresses from 16.98% to 3.78% CER through domain-specific adaptation. Overall, our results show that domain-adaptive fine-tuning enhances business-critical terminology while preserving real-time responsiveness for production voicebot deployments.

Audio reasoning is essential for machine understanding of the acoustic world. Reinforcement learning with verifiable rewards can elicit such reasoning, yet existing reward designs are complementary in their limitations: outcome-based rewards supervise only the final answer and let the model reach it without attending to the audio, whereas process-based rewards score the reasoning itself but rely on coarse, hand-crafted, and fixed criteria that neither adapt to each question nor stay grounded in the acoustic evidence. Moreover, questions differ in what they demand, with some hinging on perception and others on multi-step reasoning, and any static criterion weakens as the policy improves. Supervising the reasoning process with fine-grained, audio-grounded, and adaptive rewards is therefore crucial, yet challenging since such rewards are impractical to design by hand for every sample. To this end, we introduce AudioRubrics, a reinforcement learning framework that supervises audio reasoning with self-evolving, audio-grounded rubric rewards. AudioRubrics synthesizes per-sample rubrics from the raw waveform and, conditioned on the model's own rollouts, regenerates and reweights criteria per group, supplying a continuous learning signal that keeps targeting the current policy's weaknesses as static criteria saturate. Comprehensive evaluations across three audio reasoning benchmarks reveal that AudioRubrics substantially outperforms a wide range of open-source and training-based baselines. Furthermore, our analysis shows that the gains scale with the capability of the rubric generator and judge, and AudioRubrics converges to a stable reasoning length that avoids both degenerate collapse and unbounded growth. The improvement in audio perception further demonstrates the effectiveness of anchoring supervision in the acoustic evidence. Our project page is available at https://audiorubrics.github.io.

Sonification operates across scientific, artistic, and design contexts, yet its identity remains conceptually ambiguous: is it music, a scientific method, or a hybrid practice? This paper presents a philosophical framework that treats sonification as a coherent, transcategorical practice. Drawing on Gustavo Bueno's materialist gnoseology, we identify its operative core through systematic data-to-sound relationships that define its technical identity. Complementing this with Gilbert Simondon's philosophy of individuation, we show conceptually how sonification adapts across contexts while preserving its operative structure. By distinguishing between technical identity and pragmatic use, this framework clarifies the relationship between sonification, music, and aesthetics, providing researchers and practitioners with a philosophical framework to navigate the field's diversity while maintaining its epistemic and operational integrity.

We benchmark eleven audio classification methods: five task-aware closed-set LLMs (four Gemini models plus open-weight Kimi-Audio-7B-Instruct), four fixed-vocabulary taggers (YAMNet, PANNs, Whisper-AT, and SSLAM), a zero-shot audio-text model (CLAP), and an audio-grounded LLM (BAT). We evaluate them on a closed-set sound-source identification task over 2,242 clips spanning 23 fine-grained classes and 11 categories. Since these methods differ fundamentally in how they receive the task and how outputs are scored, we group them into four evaluation tiers rather than one leaderboard, reporting macro Precision, Recall, F1, and false-negative rate per tier. The best model, Gemini-3.1-Pro-Preview, reaches 85.6 percent category-level F1 and 56.7 percent fine-grained F1. Kimi-Audio is competitive for its size, reaching 67.5 percent category-level F1 and 32.9 percent fine-grained F1, but fails to answer 1.6 percent of samples. SSLAM and CLAP match or exceed the best closed-set model at the category level without seeing the candidate list, but fall behind at the fine-grained level. Analyzing the Gemini models' chain-of-thought across 8,968 responses, we find that response length does not predict accuracy, an apparent "holistic judgment beats detailed analysis" effect is better explained as a difficulty confound, and wrong answers are stated confidently 92 to 100 percent of the time. We report full per-class confusion matrices and metrics for all eleven methods, identify the structural error modes behind most of the accuracy loss between granularities, and give practical guidance for choosing among these method families.

Submarine fiber-optic cables instrumented with distributed acoustic sensing (DAS) provide an effective approach for large-scale monitoring of fin whales. We present an end-to-end workflow for detecting, characterizing, and localizing fin whale notes, tested on two submarine telecom cables in the Strait of Gibraltar and western Alboran Sea. The workflow applies a kurtosis-value picker adapted to narrow-band fin whale notes. Channel-wise detections are grouped into individual notes using density-based spatio-temporal clustering, cluster agglomeration, and hyperbolic fitting to reject incoherent picks. The retained clusters are characterized through temporal, spectral, and energy-related descriptors that support note-type discrimination and estimation of inter-note intervals. Relative arrival times across DAS channels are then used in a grid-search procedure to estimate candidate source locations. Evaluation against manually annotated detections from six fin whale songs yielded median pick-level precision of 0.990 and recall of 0.744, and median cluster-level precision of 0.880 and recall of 0.806. Representative applications demonstrate separation of overlapping vocalizations, characterization of type-A and type-B notes, and the inference of apparent source movement. By transforming dense DAS recordings into compact note-level bioacoustic information, the workflow provides an integrated framework for fin whale monitoring and a basis for adaptation to other synchronized acoustic receiver arrays.

Sounding Canvas turns painting into a touch-responsive multimodal installation by embedding capacitive sensors, real-time decision models, and networking inside the canvas. Touches trigger spatialised sounds that appear to emanate from the painting itself. The work embeds algorithms physically, as sensing and computation concealed behind the artwork; perceptually, through an offline visual-to-sonic mapping that aligns a painting's features with sound descriptors; and performatively, through online models that shape live interaction with visitors and with remote canvases over a network. We describe the artistic rationale and technical implementation, combining a CNN-based offline mapping that defines the sound vocabulary with two online event managers, a higher-order Markov model and an LSTM-based policy, that balance responsiveness with guided exploration. We discuss how these layers make algorithms perceptible through behaviour rather than code, how networking transforms solitary touch into distributed co-authorship, and how the system raises questions of authorship, agency, and evaluation in embedded algorithmic artworks.

Artificial intelligence offers substantial potential for acoustic monitoring of animals, from welfare assessment in precision livestock farming to wildlife conservation and ecological research, where vocalizations can indicate health, stress, and social states earlier and at lower cost than manual observation. However, recordings in these settings are obtained under uncontrolled conditions, including environmental noise, reverberation, overlapping calls, and sensors that degrade without notice. As a consequence, automated classification of animal vocalizations remains challenging, and the two dominant acoustic representations show complementary limitations: raw waveforms preserve temporal microstructure but degrade under clipping and reverberation, while log-Mel spectrograms capture harmonic organization but lose phase information and are sensitive to broadband noise. To address these challenges, we propose Uncertainty-Aware Fusion (UAF), a dual-stream framework that estimates Gaussian uncertainty for each representation and fuses them via uncertainty weighting. This mechanism assigns greater weight to the more confident representation with no reliability labels required. In a cross-species, identity-based evaluation excluding all individuals seen during training, UAF (mean pooling) achieves 59.4% accuracy / 39.7% macro F1 on the 17-class SoundWel pig vocalization benchmark and 73.1% accuracy / 71.5% macro F1 on the 3-class DogBark dataset, outperforming static-concatenation fusion by 15.7% and 20.4% relative macro F1, respectively. Ablations over four temporal aggregation strategies show that uncertainty fusion, rather than the temporal characteristics of animal calls, is the primary driver of the performance gain.

Recent audio-video generators can produce semantically plausible and apparently synchronized sound, yet may still violate the acoustic processes implied by visible events and environments. Existing benchmarks provide limited support for attributing such violations to particular acoustic processes and quantifying their severity. We introduce AcoustiTrace, a diagnostic benchmark that formalizes acoustic physical realism in audio-video generation. AcoustiTrace organizes text-to-audio-video (T2AV) and image-to-audio-video (I2AV) evaluation around the acoustic process, covering sound generation, propagation environment, and acoustic reception through eight dimensions grounded in measurable acoustic quantities. Based on these evaluation dimensions, we construct a large-scale dataset organized around acoustic mechanisms, comprising real-world audio-video recordings and acoustically annotated RGB-D observations, and use it to develop targeted prompt suites and validated evaluators. Experiments reveal that even leading generators still struggle with fundamental acoustic processes despite producing plausible sound events. Finally, we show that the diagnostics AcoustiTrace provides for specific acoustic relations can guide model refinement toward more physically faithful audio and open new directions for incorporating acoustic principles into training objectives, reward modeling, and candidate selection.

Speech and audio generation is often needed in animation dubbing, audio drama, movies, advertising, games, podcasts, and short-video production. In these scenarios, creators may need to design voices without reference recordings, control speaker styles with natural language, support acoustic scenes with environments and audio effects, and later reuse the designed voices. Therefore, it is important to support multi-speaker speech and audio generation for both instruct and zero-shot tasks. The instruct task requires a caption of the environment, speaker styles, and fine-grained content, while the zero-shot task uses reference audio together with the same fine-grained content. We address these tasks from both the data and model sides. First, we propose SwanData-Caption, which cleans raw speech and audio data, adds targeted synthetic coverage, and annotates diverse and accurate multi-level captions. Then, we propose SwanTale, a multi-speaker expressive speech and audio generation model that supports both zero-shot and instruct tasks. We introduce SwanVAE to support high-quality multi-audio-modality generation. Then, we adopt reward-conditioned quality control and Engram conditioning, along with Unified MoE for multi-task and multi-audio-modality modeling. In addition, we use curriculum learning and GRPO post-training to let the model progressively learn and strengthen its capabilities. Experimental results show that SwanTale leads on multiple key zero-shot and instruct metrics, achieves the best expressiveness scores in both tasks, and supports complex instruct generation involving multi-speaker speech and audio. Demos can be found at https://swanaigc.github.io/#swantale.

MIDI-to-Music system renders the melody and rhythm of a target MIDI sequence into musical segment while cloning instrument timbre from a prompt recording. Existing systems typically adopt one of two distinct paradigms: conditional generation with prompt audio alone, which remains applicable when aligned prompt MIDI is unavailable, and In-Context Learning with paired prompt audio and MIDI, which exploits cross-modal alignment for stronger control on MIDI following and timbre similarity. We introduce P-MUSE, an instrumental MIDI-to-Music framework that unifies both paradigms via a multi-stage Curriculum-Learning supporting prompt-MIDI-optional inputs. P-MUSE further unifies music generation and local editing through a shared fill-in-the-middle formulation. Grounded in theoretical analysis and empirical study, we propose a phase-aware classifier-free guidance scheduling principle for Transcription-to-Audio systems, alongside a Tail-Drop strategy. Finally, to advance research in this field, we establish the first comprehensive benchmark, covering various prompt modes, generation/editing tasks, and four representative instruments: piano, guitar, bass, and drums. Demos are available at https://p-muse.github.io/.

This paper describes the participation of team "Go-To-Germany" in the ImageCLEF 2026 Audio Deepfake Detection and Generation task. Our detection system, built on a four-backbone self-supervised learning (SSL) ensemble combining WavLM-Large, Wav2Vec2-XLS-R-300M, ECAPA-TDNN, and x-vector representations, achieved a final score of 0.9522 on the official ImageCLEF 2026 evaluation, with perfect accuracy (1.0000) on participant-generated deepfakes and 0.8875 on the held-out organizer ground-truth real data. For the Generation sub-task, our official team submission, an F5-TTS v1 baseline processed with a uniform reverberation pass and submitted as a deliberate anti-forensic probe, ranked first with a final score of 0.4304 (word error rate (WER) 4.99%, character error rate (CER) 2.07%); details of our four-model program (GLM-TTS, F5-TTS, XTTS v2, CosyVoice3), from which the official entry was drawn, appear in the paper. We present a cross-track analysis revealing a pronounced asymmetry: our detection system identifies 100% of participant-generated deepfakes, while our official generation entry, despite ranking first in the Audio Generation sub-task and evading 61.4% and 56.2% of participant and organizer detectors, attains a Final Score of 0.4304 against 0.9522 on the Detection side. We further report falsification-based ablation experiments (LOSO 56-speaker cross-validation, three-region backbone geometry, bootstrap confidence intervals, and PCA analysis) that motivate our architectural-insurance hypothesis for multi-backbone SSL ensembling. We complement these results with five cross-track insights and five pre-registered falsification experiments connecting generation-side evasion to detection-side design decisions, and we openly report an 11.25% false-positive gap on held-out organizer real recordings as the principal open challenge for deployment.

Scoring open-ended music analysis responses is time-consuming and requires nuanced judgments of harmonic knowledge and formal understanding. This study evaluates the validity and repeatability of GPT-4o-mini for rubric-based scoring of music analysis essays, using teacher mean scores as the benchmark. A dataset of 300 university-level student responses was scored by teachers on four dimensions: Harmony, Form, Reasoning, and Terminology. GPT-4o-mini scored the same responses using three prompting strategies: few-shot prompting with chain-of-thought reasoning (Fs+CoT), retrieval-augmented generation (RAG), and self-consistency based on five internal generations per administration (SC). Each strategy was administered three times with the model, prompt, rubric, and response held constant. Single-pass scores represented an operational scoring condition, whereas median aggregation across three runs was used to examine robustness. Agreement with teacher mean scores was evaluated using correlation, intraclass correlation, Krippendorff's alpha, quadratic weighted kappa, and scoring error indices. Fs+CoT showed the strongest agreement with teacher mean scores in both single-pass scoring and median aggregation. RAG showed systematic over-scoring, whereas SC produced highly repeatable scores but weaker individual-level agreement. Dimension-level analyses showed that scoring performance varied across rubric components, with Terminology generally showing weaker agreement than Reasoning. These findings indicate that GPT-4o-mini can generate stable scores for complex music analysis responses, but prompting strategies produce distinct scoring profiles. Operational use therefore requires strategy-specific calibration, dimension-level validation, and continued human oversight.

Token compression in OmniLLMs is typically posed as a single saliency-ranking problem: score each multimodal token, keep the top-K. We argue this abstraction is mis-specified. The same attention score simultaneously decides two things: how much retained capacity each modality receives, and which tokens within a modality are kept. A shared top-K rule therefore inherits this audio-favoring allocation prior, spending retained capacity on audio before video tokens have a chance to compete. We propose Macer, a training-free compressor that first assigns explicit audio and video budgets, then performs allocation-normalized ranking within each modality at modality-specific shallow layers. Macer significantly reduces token cost while preserving accuracy across audio-grounded, audio–video joint, visual-dominant, and video-centric benchmarks. At 25 % retention, Macer preserves 98.7 % of full-token performance on Qwen2.5-Omni-7B and 97.3 % on Qwen2.5-Omni-3B. On Qwen2.5-Omni-7B, this 25 % setting reaches OmniZip-level performance at 45 % retention while using lower FLOPs. On OmniVinci-9B, the same allocation-before-ranking principle improves over shared top-K ranking by up to 12.9 points.

EEG-guided target speaker extraction is challenging under in-trial auditory attention switching, where neural noise and intrinsic latency can delay or destabilize attention tracking. Conventional methods struggle with dynamic switches and often cause discontinuities at switching points. Therefore, we propose SAGE, a switch-aware EEG-guided soft gating framework that treats in-trial switching as dynamic selection. SAGE generates two candidate speech streams with a robust separator and uses an EEG-guided switch-aware gating module to produce smooth fusion weights and suppress transition artifacts. We further integrate latency-compensated alignment and an uncertainty-driven conservative strategy to handle latency discrepancies and fluctuating EEG reliability. SAGE outperforms baselines, achieving 8.67 dB SI-SDR and 88.24% STOI while reducing average switching latency to 2.04 s. By coupling neural decoding with speech separation, it enables robust target extraction in dynamic scenarios.

Achieving robust EEG-based auditory attention switch decoding (AASD) is crucial for intelligent hearing aids. However, its application is limited as EEG non-stationarity complicates sequential decision-making, and insufficient control of potential confounding factors may overestimate performance. Therefore, we propose a state-guided adaptive decision (SGAD) framework that infers attention transition states via causal state detection and dynamically modulates temporal smoothing through state-guided adaptive gating. We further introduce six hierarchical evaluation protocols to assess generalization across audio, speaker, and subject dimensions. Experimental results show that SGAD improves decoding accuracy and stability while maintaining low response latency across evaluation scenarios. Performance variations across protocols further suggest data partition-related biases. Together, these findings advance robust AASD for neuro-steered hearing applications.

Scaling the corpus is the default remedy when a contrastive representation lacks an attribute. We report a case where it does nothing, and identify what does: adding a lexical-speech round to a frozen-base multimodal embedding model raises zero-shot keyword spotting by 76 points while reducing speech-emotion recognition by 14. The loss is not a capacity limit: fine-tuning on 7,442 clips from a prosody-controlled corpus recovers emotion past its pre-speech level at a five-point keyword cost. Nor is it data volume: 29,428 mined clips whose captions explicitly name emotions, at matched exposure, move emotion by -0.0007. The difference is structural: a contrastive objective encodes an attribute only when the in-batch negatives cannot be separated without it; the controlled corpus holds sentence content fixed, so prosody is the only separating signal, whereas mined captions name emotion yet remain separable by scene content. Intervention on the same audio confirms causality: raising caption similarity does not recover emotion, but collapsing caption diversity so that emotion becomes the only separating axis recovers it by 8.9 points across three seeds, with a smaller, same-signed gain on a non-acted corpus, while keyword accuracy trades back. Corpus structure, not size or caption vocabulary, controls what a contrastive audio embedding encodes.