DNSMOS-C: Improving End-to-end Speech Quality Models via Contrastive Learning

We introduce DNSMOS-C, a compact end-to-end speech quality assessment model that extends the DNSMOS Pro framework by integrating a MOS-guided triplet-based contrastive loss. Applied directly to the intermediate embeddings, this contrastive supervision encourages the latent space to be better organized with respect to perceptual quality while preserving the simplicity and efficiency of DNSMOS Pro. Unlike prior methods that depend on large pre-trained self-supervised learning (SSL) encoders and multi-stage training, DNSMOS-C jointly learns speech representations and MOS regression within a single, unified framework. Experiments on multiple datasets show that DNSMOS-C consistently improves correlation metrics over DNSMOS Pro and achieves better generalization on challenging out-of-domain test sets. Furthermore, latent space analyses indicate that our approach learns representations that exhibit an emergent low-dimensional quality ordering, which enhances interpretability and improves training stability. These findings demonstrate that MOS-guided contrastive learning enables more robust and accurate quality predictions without incurring additional computational overhead.

Generative Modeling for Physiological Signals

Physiological signals support clinical diagnosis, health monitoring, rehabilitation, wearable sensing, and human--machine interaction. However, their applications are often constrained by limited labeled data, class imbalance, noisy or incomplete recordings, heterogeneous acquisition settings, and privacy restrictions. Generative modeling has therefore attracted increasing attention as a means of addressing some of these barriers. Recent studies have used generative models to augment scarce datasets, restore degraded recordings, translate between modalities, and synthesize conditional physiological waveforms. This review summarizes recent work on generative modeling for cardiovascular, neural, muscular, peripheral, and specialized physiological signals. Major model families are covered, including generative adversarial networks (GANs), autoencoders and variational autoencoders (AEs/VAEs), diffusion models, autoregressive sequence models, and hybrid architectures. In addition, it organizes existing evaluation practices into a hierarchical framework spanning signal-level similarity, dataset-level distribution, physiological validity, task-oriented utility, and assessments of generalization and robustness. By linking signal-specific constraints, generative roles, model families, and evaluation evidence, this review provides structured guidance for the future use and evaluation of generative models in physiological-signal research.

Diffusion-Based Heart Sound Generation: Evaluation with Physiological Signal Metrics, Classifiers, and Expert Listening

Publicly available phonocardiogram (PCG) datasets remain limited in size and pathological diversity, constraining both auscultation training and the generalisation of automated heart-sound classifiers. A class-conditional diffusion model for PCG generation is developed in the log-mel domain and synthetic fidelity is assessed using complementary (i) physiology-inspired plausibility metrics, (ii) downstream label-consistency evaluation, and (iii) expert listening. Experiments use the Phy-sioNet/Computing in Cardiology Challenge 2016 dataset (3240 recordings) with recording-level splits. After preprocessing and quality control, 16,749 non-overlapping 4 s clips are mapped to a normalised 1 x 128 x 128 log-mel representation to train a conditional 2D U-Net denoiser with classifier-free guidance. Signal-level plausibility is quantified on reconstructed waveforms using three lightweight metrics: an envelope-autocorrelation rhythm score, an amplitude-based explosion score, and the dominant cycle lag. Synthetic clips preserve similar dominant cycle durations but exhibit reduced envelope periodicity and increased transient burstiness relative to real clips. For downstream evaluation, a ResNet-50 classifier achieves 92.24% accuracy on the held-out real test set and 82.8% accuracy on class-balanced synthetic batches, indicating that generated signals retain discriminative structure relevant to normal/abnormal classification. In a pilot expert listening study (60 clips, two clinicians), most synthetic clips are judged as heart-sound-like, while abnormality sensitivity is low for both real and synthetic 4 s excerpts. Overall, the results provide a practical baseline for diffusion-based PCG generation while highlighting remaining challenges in retaining abnormal acoustic cues and reducing reconstruction-induced artefacts.

SA-SSL-MOS: Self-supervised Learning MOS Prediction with Spectral Augmentation for Generalized Multi-Rate Speech Assessment

Designing a speech quality assessment (SQA) system for estimating mean-opinion-score (MOS) of multi-rate speech with varying sampling frequency (16-48 kHz) is a challenging task. The challenge arises due to the limited availability of a MOS-labeled training dataset comprising multi-rate speech samples. While self-supervised learning (SSL) models have been widely adopted in SQA to boost performance, a key limitation is that they are pretrained on 16 kHz speech and therefore discard high-frequency information present in higher sampling rates. To address this issue, we propose a spectrogram-augmented SSL method that incorporates high-frequency features (up to 48 kHz sampling rate) through a parallel-branch architecture. We further introduce a two-step training scheme: the model is first pre-trained on a large 48 kHz dataset and then fine-tuned on a smaller multi-rate dataset. Experimental results show that leveraging high-frequency information overlooked by SSL features is crucial for accurate multi-rate SQA, and that the proposed two-step training substantially improves generalization when multi-rate data is limited.

DNS: Data-driven Nonlinear Smoother for Complex Model-free Process

We propose data-driven nonlinear smoother (DNS) to estimate a hidden state sequence of a complex dynamical process from a noisy, linear measurement sequence. The dynamical process is model-free, that is, we do not have any knowledge of the nonlinear dynamics of the complex process. There is no state-transition model (STM) of the process available. The proposed DNS uses a recurrent architecture that helps to provide a closed-form posterior of the hidden state sequence given the measurement sequence. DNS learns in an unsupervised manner, meaning the training dataset consists of only measurement data and no state data. We demonstrate DNS using simulations for smoothing of several stochastic dynamical processes, including a benchmark Lorenz system. Experimental results show that the DNS is significantly better than a deep Kalman smoother (DKS) and an iterative data-driven nonlinear state estimation (iDANSE) smoother.

VSE: Variational state estimation of complex model-free process

We design a variational state estimation (VSE) method that provides a closed-form Gaussian posterior of an underlying complex dynamical process from (noisy) nonlinear measurements. The complex process is model-free. That is, we do not have a suitable physics-based model characterizing the temporal evolution of the process state. The closed-form Gaussian posterior is provided by a recurrent neural network (RNN). The use of RNN is computationally simple in the inference phase. For learning the RNN, an additional RNN is used in the learning phase. Both RNNs help each other learn better based on variational inference principles. The VSE is demonstrated for a tracking application - state estimation of a stochastic Lorenz system (a benchmark process) using a 2-D camera measurement model. The VSE is shown to be competitive against a particle filter that knows the Lorenz system model and a recently proposed data-driven state estimation method that does not know the Lorenz system model.

pDANSE: Particle-based Data-driven Nonlinear State Estimation from Nonlinear Measurements

We consider the problem of designing a data-driven nonlinear state estimation (DANSE) method that uses (noisy) nonlinear measurements of a process whose underlying state transition model (STM) is unknown. Such a process is referred to as a model-free process. A recurrent neural network (RNN) provides parameters of a Gaussian prior that characterize the state of the model-free process, using all previous measurements at a given time point. In the case of DANSE, the measurement system was linear, leading to a closed-form solution for the state posterior. However, the presence of a nonlinear measurement system renders a closed-form solution infeasible. Instead, the second-order statistics of the state posterior are computed using the nonlinear measurements observed at the time point. We address the nonlinear measurements using a reparameterization trick-based particle sampling approach, and estimate the second-order statistics of the state posterior. The proposed method is referred to as particle-based DANSE (pDANSE). The RNN of pDANSE uses sequential measurements efficiently and avoids the use of computationally intensive sequential Monte-Carlo (SMC) and/or ancestral sampling. We describe the semi-supervised learning method for pDANSE, which transitions to unsupervised learning in the absence of labeled data. Using a stochastic Lorenz-$63$ system as a benchmark process, we experimentally demonstrate the state estimation performance for four nonlinear measurement systems. We explore cubic nonlinearity and a camera-model nonlinearity where unsupervised learning is used; then we explore half-wave rectification nonlinearity and Cartesian-to-spherical nonlinearity where semi-supervised learning is used. The performance of state estimation is shown to be competitive vis-\`a-vis particle filters that have complete knowledge of the STM of the Lorenz-$63$ system.

Selection of Layers from Self-supervised Learning Models for Predicting Mean-Opinion-Score of Speech

Self-supervised learning (SSL) models like Wav2Vec2, HuBERT, and WavLM have been widely used in speech processing. These transformer-based models consist of multiple layers, each capturing different levels of representation. While prior studies explored their layer-wise representations for efficiency and performance, speech quality assessment (SQA) models predominantly rely on last-layer features, leaving intermediate layers underexamined. In this work, we systematically evaluate different layers of multiple SSL models for predicting mean-opinion-score (MOS). Features from each layer are fed into a lightweight regression network to assess effectiveness. Our experiments consistently show early-layers features outperform or match those from the last layer, leading to significant improvements over conventional approaches and state-of-the-art MOS prediction models. These findings highlight the advantages of early-layer selection, offering enhanced performance and reduced system complexity.

Leveraging LLMs for Scalable Non-intrusive Speech Quality Assessment

Non-intrusive speech quality assessment (SQA) systems suffer from limited training data and costly human annotations, hindering their generalization to real-time conferencing calls. In this work, we propose leveraging large language models (LLMs) as pseudo-raters for speech quality to address these data bottlenecks. We construct LibriAugmented, a dataset consisting of 101,129 speech clips with simulated degradations labeled by a fine-tuned auditory LLM (Vicuna-7b-v1.5). We compare three training strategies: using human-labeled data, using LLM-labeled data, and a two-stage approach (pretraining on LLM labels, then fine-tuning on human labels), using both DNSMOS Pro and DeePMOS. We test on several datasets across languages and quality degradations. While LLM-labeled training yields mixed results compared to human-labeled training, we provide empirical evidence that the two-stage approach improves the generalization performance (e.g., DNSMOS Pro achieves 0.63 vs. 0.55 PCC on NISQA_TEST_LIVETALK and 0.73 vs. 0.65 PCC on Tencent with reverb). Our findings demonstrate the potential of using LLMs as scalable pseudo-raters for speech quality assessment, offering a cost-effective solution to the data limitation problem.

Multivariate Probabilistic Assessment of Speech Quality

The mean opinion score (MOS) is a standard metric for assessing speech quality, but its singular focus fails to identify specific distortions when low scores are observed. The NISQA dataset addresses this limitation by providing ratings across four additional dimensions: noisiness, coloration, discontinuity, and loudness, alongside MOS. In this paper, we extend the explored univariate MOS estimation to a multivariate framework by modeling these dimensions jointly using a multivariate Gaussian distribution. Our approach utilizes Cholesky decomposition to predict covariances without imposing restrictive assumptions and extends probabilistic affine transformations to a multivariate context. Experimental results show that our model performs on par with state-of-the-art methods in point estimation, while uniquely providing uncertainty and correlation estimates across speech quality dimensions. This enables better diagnosis of poor speech quality and informs targeted improvements.