arXiv subjects

Yunyun Wang

Publications and source records attributed to Yunyun Wang.

12 recordsLinked to original sources

arXiv2026

ParA-LLM: A Unified Approach to Paralinguistic and Acoustic Speech Understanding

Recent advances in Audio LLMs have achieved human-level speech recognition, yet existing systems struggle to capture paralinguistic aspects such as speaker traits, expressive variations, and environmental acoustic conditions. To address this, we design a framework of 22 paralinguistic characteristics and create a dataset of over 1.2M Audio-QA pairs. We develop ParA-LLM, trained with a two-stage curriculum: first on single-attribute questions to build foundational knowledge, then on multi-attribute questions for joint reasoning over speaker and acoustic characteristics. We also release ParA-Bench, a benchmark of 6,000 multiple-choice questions across speaker-speech, acoustic, and mixed categories, where frontier models like GPT-4o-Audio achieve only 36% accuracy. ParA-LLM surpasses state-of-the-art Audio LLMs like GPT-4o-Audio by 7.5% on ParA-Bench, with additional gains of 1.13% on MMAU-Pro Speech and 7.49% on MMAR Speech.

cs.SD↗

arXiv2026

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

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

eess.AS↗

arXiv2026

Do Speech Emphasis Models Generalize across Languages and Emotions?

Prosodic emphasis varies across languages, emotions, and speaking styles, yet existing emphasis detection models are largely trained and evaluated on monolingual neutral read speech. We introduce MMEE (Multilingual Multi-Emotion Emphasis), a corpus of 10,000 professionally recorded expressive utterances (14.13 hours) across 7 languages and 34 emotion/style categories, with three-level perceptual labels (10 annotations per sample). We benchmark two state-of-the-art architectures under monolingual, cross-lingual, multilingual, cross-emotion, cross-dataset, and data-scale settings. Monolingual models show limited zero-shot transfer, degrading across typologically distant languages, while multilingual training substantially improves robustness. Models transfer robustly between high- and low-arousal emotions; bidirectional transfer between synthetic and perceptual benchmarks suggests shared prosodic structure; and performance stays robust even at smaller training scales.

cs.CL↗

arXiv2026

MedCoRAG: Interpretable Hepatology Diagnosis via Hybrid Evidence Retrieval and Multispecialty Consensus

Diagnosing hepatic diseases accurately and interpretably is critical, yet it remains challenging in real-world clinical settings. Existing AI approaches for clinical diagnosis often lack transparency, structured reasoning, and deployability. Recent efforts have leveraged large language models (LLMs), retrieval-augmented generation (RAG), and multi-agent collaboration. However, these approaches typically retrieve evidence from a single source and fail to support iterative, role-specialized deliberation grounded in structured clinical data. To address this, we propose MedCoRAG (i.e., Medical Collaborative RAG), an end-to-end framework that generates diagnostic hypotheses from standardized abnormal findings and constructs a patient-specific evidence package by jointly retrieving and pruning UMLS knowledge graph paths and clinical guidelines. It then performs Multi-Agent Collaborative Reasoning: a Router Agent dynamically dispatches Specialist Agents based on case complexity; these agents iteratively reason over the evidence and trigger targeted re-retrievals when needed, while a Generalist Agent synthesizes all deliberations into a traceable consensus diagnosis that emulates multidisciplinary consultation. Experimental results on hepatic disease cases from MIMIC-IV show that MedCoRAG outperforms existing methods and closed-source models in both diagnostic performance and reasoning interpretability.

cs.AI↗

arXiv2026

Inflation Target at Risk: A Time-varying Parameter Distributional Regression

Inflation exhibits state-dependent, skewed, and fat-tailed dynamics that make risk a central concern for monetary policy. Accordingly, inflation risks are distributional and cannot be fully captured by mean-based models. We propose a flexible time-varying parameter distributional regression model that estimates the full conditional distribution of inflation, allowing macroeconomic drivers to have nonlinear and asymmetric effects across the distribution. Applied to U.S. inflation, the model captures major shifts in tail-risk probabilities. Analysis of risk drivers shows that deflationary pressures arise primarily from demand-side weakness and inflation persistence, whereas upside risks are driven mainly by supply-side shocks, particularly energy price inflation. Examining the impact of key drivers further reveals that the unemployment-inflation relationship weakens in the distributional tails. Energy price shocks, by contrast, have little effect on deflation risk but exhibit strongly time-varying and asymmetric effects on high-inflation risk.

econ.EM↗

arXiv2025

The Finer the Better: Towards Granular-aware Open-set Domain Generalization

Open-Set Domain Generalization (OSDG) tackles the realistic scenario where deployed models encounter both domain shifts and novel object categories. Despite impressive progress with vision-language models like CLIP, existing methods still fall into the dilemma between structural risk of known-classes and open-space risk from unknown-classes, and easily suffers from over-confidence, especially when distinguishing ``hard unknowns" that share fine-grained visual similarities with known classes. To this end, we propose a Semantic-enhanced CLIP (SeeCLIP) framework that explicitly addresses this dilemma through fine-grained semantic enhancement. In SeeCLIP, we propose a semantic-aware prompt enhancement module to decompose images into discriminative semantic tokens, enabling nuanced vision-language alignment beyond coarse category labels. To position unknown prompts effectively, we introduce duplex contrastive learning with complementary objectives, that is, repulsion to maintain separability from known classes, and cohesion to preserve semantic proximity. Further, our semantic-guided diffusion module synthesizes pseudo-unknowns by perturbing extracted semantic tokens, generating challenging samples that are visually similar to known classes yet exhibit key local differences. These hard negatives force the model to learn finer decision boundaries. Extensive experiments across five benchmarks demonstrate consistent improvements of 3% accuracy and 5% H-score over state-of-the-art methods.

cs.CV↗

arXiv2025

Everyone-Can-Sing: Zero-Shot Singing Voice Synthesis and Conversion with Speech Reference

We propose a unified framework for Singing Voice Synthesis (SVS) and Conversion (SVC), addressing the limitations of existing approaches in cross-domain SVS/SVC, poor output musicality, and scarcity of singing data. Our framework enables control over multiple aspects, including language content based on lyrics, performance attributes based on a musical score, singing style and vocal techniques based on a selector, and voice identity based on a speech sample. The proposed zero-shot learning paradigm consists of one SVS model and two SVC models, utilizing pre-trained content embeddings and a diffusion-based generator. The proposed framework is also trained on mixed datasets comprising both singing and speech audio, allowing singing voice cloning based on speech reference. Experiments show substantial improvements in timbre similarity and musicality over state-of-the-art baselines, providing insights into other low-data music tasks such as instrumental style transfer. Examples can be found at: everyone-can-sing.github.io.

cs.SD↗

arXiv2023

Bivariate Distribution Regression with Application to Insurance Data

Understanding variable dependence, particularly eliciting their statistical properties given a set of covariates, provides the mathematical foundation in practical operations management such as risk analysis and decision-making given observed circumstances. This article presents an estimation method for modeling the conditional joint distribution of bivariate outcomes based on the distribution regression and factorization methods. This method is considered semiparametric in that it allows for flexible modeling of both the marginal and joint distributions conditional on covariates without imposing global parametric assumptions across the entire distribution. In contrast to existing parametric approaches, our method can accommodate discrete, continuous, or mixed variables, and provides a simple yet effective way to capture distributional dependence structures between bivariate outcomes and covariates. Various simulation results confirm that our method can perform similarly or better in finite samples compared to the alternative methods. In an application to the study of a motor third-party liability insurance portfolio, the proposed method effectively estimates risk measures such as the conditional Value-at-Risk and Expected Shortfall. This result suggests that this semiparametric approach can serve as an alternative in insurance risk management.

stat.ME↗

arXiv2023

Distributional Vector Autoregression: Eliciting Macro and Financial Dependence

Vector autoregression is an essential tool in empirical macroeconomics and finance for understanding the dynamic interdependencies among multivariate time series. In this study, we expand the scope of vector autoregression by incorporating a multivariate distributional regression framework and introducing a distributional impulse response function, providing a comprehensive view of dynamic heterogeneity. We propose a straightforward yet flexible estimation method and establish its asymptotic properties under weak dependence assumptions. Our empirical analysis examines the conditional joint distribution of GDP growth and financial conditions in the United States, with a focus on the global financial crisis. Our results show that tight financial conditions lead to a multimodal conditional joint distribution of GDP growth and financial conditions, and easing financial conditions significantly impacts long-term GDP growth, while improving the GDP growth during the global financial crisis has limited effects on financial conditions.

econ.EM↗

arXiv2022

Dual-Correction Adaptation Network for Noisy Knowledge Transfer

Previous unsupervised domain adaptation (UDA) methods aim to promote target learning via a single-directional knowledge transfer from label-rich source domain to unlabeled target domain, while its reverse adaption from target to source has not jointly been considered yet so far. In fact, in some real teaching practice, a teacher helps students learn while also gets promotion from students to some extent, which inspires us to explore a dual-directional knowledge transfer between domains, and thus propose a Dual-Correction Adaptation Network (DualCAN) in this paper. However, due to the asymmetrical label knowledge across domains, transfer from unlabeled target to labeled source poses a more difficult challenge than the common source-to-target counterpart. First, the target pseudo-labels predicted by source commonly involve noises due to model bias, hence in the reverse adaptation, they may hurt the source performance and bring a negative target-to-source transfer. Secondly, source domain usually contains innate noises, which will inevitably aggravate the target noises, leading to noise amplification across domains. To this end, we further introduce a Noise Identification and Correction (NIC) module to correct and recycle noises in both domains. To our best knowledge, this is the first naive attempt of dual-directional adaptation for noisy UDA, and naturally applicable to noise-free UDA. A theory justification is given to state the rationality of our intuition. Empirical results confirm the effectiveness of DualCAN with remarkable performance gains over state-of-the-arts, particularly for extreme noisy tasks (e.g., ~+ 15% on Pw->Pr and Pr->Rw of Office-Home).

cs.CV↗

arXiv2022

Towards Adaptive Unknown Authentication for Universal Domain Adaptation by Classifier Paradox

Universal domain adaptation (UniDA) is a general unsupervised domain adaptation setting, which addresses both domain and label shifts in adaptation. Its main challenge lies in how to identify target samples in unshared or unknown classes. Previous methods commonly strive to depict sample "confidence" along with a threshold for rejecting unknowns, and align feature distributions of shared classes across domains. However, it is still hard to pre-specify a "confidence" criterion and threshold which are adaptive to various real tasks, and a mis-prediction of unknowns further incurs misalignment of features in shared classes. In this paper, we propose a new UniDA method with adaptive Unknown Authentication by Classifier Paradox (UACP), considering that samples with paradoxical predictions are probably unknowns belonging to none of the source classes. In UACP, a composite classifier is jointly designed with two types of predictors. That is, a multi-class (MC) predictor classifies samples to one of the multiple source classes, while a binary one-vs-all (OVA) predictor further verifies the prediction by MC predictor. Samples with verification failure or paradox are identified as unknowns. Further, instead of feature alignment for shared classes, implicit domain alignment is conducted in output space such that samples across domains share the same decision boundary, though with feature discrepancy. Empirical results validate UACP under both open-set and universal UDA settings.

cs.CV↗

arXiv2022

Pre-Training on Dynamic Graph Neural Networks

The pre-training on the graph neural network model can learn the general features of large-scale networks or networks of the same type by self-supervised methods, which allows the model to work even when node labels are missing. However, the existing pre-training methods do not take network evolution into consideration. This paper proposes a pre-training method on dynamic graph neural networks (PT-DGNN), which uses dynamic attributed graph generation tasks to simultaneously learn the structure, semantics, and evolution features of the graph. The method includes two steps: 1) dynamic sub-graph sampling, and 2) pre-training with dynamic attributed graph generation task. Comparative experiments on three realistic dynamic network datasets show that the proposed method achieves the best results on the link prediction fine-tuning task.

cs.LG↗