Papers

Accurate and responsive myoelectric prosthesis control typically relies on complex, dense multi-sensor arrays, which limits consumer accessibility. This paper presents a novel, data-efficient deep learning framework designed to achieve precise and accurate control using minimal sensor hardware. Leveraging an external dataset of 8 subjects, our approach implements a hybrid Transformer optimized for sparse, two-channel surface electromyography (sEMG). Unlike standard architectures that use fixed positional encodings, we integrate Time2Vec learnable temporal embeddings to capture the stochastic temporal warping inherent in biological signals. Furthermore, we employ a normalized additive fusion strategy that aligns the latent distributions of spatial and temporal features, preventing the destructive interference common in standard implementations. A two-stage curriculum learning protocol is utilized to ensure robust feature extraction despite data scarcity. The proposed architecture achieves a state-of-the-art multi-subject F1-score of 95.7% $\pm$ 0.20% for a 10-class movement set, statistically outperforming both a standard Transformer with fixed encodings and a recurrent CNN-LSTM model. Architectural optimization reveals that a balanced allocation of model capacity between spatial and temporal dimensions yields the highest stability. Furthermore, while direct transfer to a new unseen subject led to poor accuracy due to domain shifts, a rapid calibration protocol utilizing only two trials per gesture recovered performance from 21.0% $\pm$ 2.98% to 96.9% $\pm$ 0.52%. By validating that high-fidelity temporal embeddings can compensate for low spatial resolution, this work challenges the necessity of high-density sensing. The proposed framework offers a robust, cost-effective blueprint for next-generation prosthetic interfaces capable of rapid personalization.

Data collaboration between municipal authorities (MA) and mobility providers (MPs) has brought tremendous benefits to transportation systems in the era of big data. Engaging in collaboration can improve the service operations (e.g., reduced delay) of these data owners, however, it can also raise privacy concerns and discourage data-sharing willingness. Specifically, data owners may be concerned that the shared data may leak sensitive information about their customers' mobility patterns or business secrets, resulting in the failure of collaboration. This paper investigates how privacy-preserving mechanisms can foster data collaboration in such settings. We propose a game-theoretic framework to investigate data-sharing among transportation stakeholders, especially considering perturbation-based privacy-preserving mechanisms. Numerical studies demonstrate that lower data quality expectations can incentivize voluntary data sharing, improving transport-related welfare for both MAs and MPs. Our findings provide actionable insights for policymakers and system designers on how privacy-preserving technologies can help bridge data silos and promote collaborative, privacy-aware transportation systems.

Transmission line (TL) models implemented in the time domain can efficiently simulate basilar-membrane (BM) displacement in response to transient or non-stationary sounds. By design, a TL model is well-suited for an one-dimensional (1-D) characterization of the traveling wave, but the real configuration of the cochlea also introduces higher-dimensional effects. Such effects include the focusing of the pressure around the BM and transverse viscous damping, both of which are magnified in the short-wave region. The two effects depend on the wavelength and are more readily expressed in the frequency domain. In this paper, we introduce a numerical correction for the BM admittance to account for 2-D effects in the time domain using autoregressive filtering and regression techniques. The correction was required for the implementation of a TL model tailored to the gerbil cochlear physiology. The model, which includes instantaneous nonlinearities in the form of variable damping, initially presented insufficient compression with increasing sound levels. This limitation was explained by the strong coupling between gain and frequency selectivity assumed in the 1-D nonlinear TL model, whereas cochlear frequency selectivity shows only a moderate dependence on sound level in small mammals. The correction factor was implemented in the gerbil model and made level-dependent using a feedback loop. The updated model achieved some decoupling between frequency selectivity and gain, providing 5 dB of additional gain and extending the range of sound levels of the compressive regime by 10 dB. We discuss the relevance of this work through two key features: the integration of both analytical and regression methods for characterizing BM admittance, and the combination of instantaneous and non-instantaneous nonlinearities.

Spoofing-aware speaker verification (SASV) jointly addresses automatic speaker verification and spoofing countermeasures to improve robustness against adversarial attacks. In this paper, we investigate our recently proposed modular SASV framework that enables effective reuse of publicly available ASV and CM systems through non-linear fusion, explicitly modeling their interaction, and optimization with an operating-condition-dependent trainable a-DCF loss. The framework is evaluated using ECAPA-TDNN and ReDimNet as ASV embedding extractors and SSL-AASIST as the CM model, with experiments conducted both with and without fine-tuning on the WildSpoof SASV training data. Results show that the best performance is achieved by combining ReDimNet-based ASV embeddings with fine-tuned SSL-AASIST representations, yielding an a-DCF of 0.0515 on the progress evaluation set and 0.2163 on the final evaluation set.

Current deep learning models for Multispectral and Hyperspectral Image Fusion (MS/HS fusion) are typically designed for fixed spectral bands and spatial scales, which limits their transferability across diverse sensors. To address this, we propose SSA, a universal framework for MS/HS fusion with spectral-band and fusion-scale agnosticism. Specifically, we introduce Matryoshka Kernel (MK), a novel operator that enables a single model to adapt to arbitrary numbers of spectral channels. Meanwhile, we build SSA upon an Implicit Neural Representation (INR) backbone that models the HS signal as a continuous function, enabling reconstruction at arbitrary spatial resolutions. Together, these two forms of agnosticism enable a single MS/HS fusion model that generalizes effectively to unseen sensors and spatial scales. Extensive experiments demonstrate that our single model achieves state-of-the-art performance while generalizing well to unseen sensors and scales, paving the way toward future HS foundation models.

Stacked intelligent metasurfaces (SIMs), composed of multiple layers of reconfigurable transmissive metasurfaces, are gaining prominence as a transformative technology for future wireless communication security. This paper investigates the integration of SIM into multi-user multiple-input multiple-output (MIMO) systems to enhance physical layer security. A novel system architecture is proposed, wherein each base station (BS) antenna transmits a dedicated single-user stream, while a multi-layer SIM executes wave-based beamforming in the electromagnetic domain, thereby avoiding the need for complex baseband digital precoding and significantly reducing hardware overhead. To maximize the weighted sum secrecy rate (WSSR), we formulate a joint precoding optimization problem over BS power allocation and SIM phase shifts, which is high-dimensional and non-convex due to the complexity of the objective function and the coupling among optimization variables. To address this, we propose a manifold-enhanced heterogeneous multi-agent continual learning (MHACL) framework that incorporates gradient representation and dual-scale policy optimization to achieve robust performance in dynamic environments with high demands for secure communication. Furthermore, we develop SIM-MHACL (SIMHACL), a low-complexity learning template that embeds phase coordination into a product manifold structure, reducing the exponential search space to linear complexity while maintaining physical feasibility. Simulation results validate that the proposed framework achieves millisecond-level per-iteratio ntraining in SIM-assisted systems, significantly outperforming various baseline schemes, with SIMHACL achieving comparable WSSR to MHACL while reducing computation time by 30%.

Angle-resolved channel sounding using antenna arrays or mechanically steered high-gain antennas is widely employed at millimeter-wave and terahertz bands. To extract antenna-independent large-scale channel parameters such as path loss, delay spread, and angular spread, the radiation-pattern effects embedded in the measured responses must be properly compensated. This paper revisits the technical challenges of path loss/path gain calculation from angle-resolved wideband measurements, with emphasis on angular-domain power integration where the scan beams are inherently non-orthogonal and simple power summation leads to biased isotropic-equivalent power estimates. We first formulate the synthesized-isotropic narrowband power in a unified matrix form and introduce a beam-accumulation correction factor, including an offset-averaged variant to mitigate scalloping due to off-grid angles. The proposed framework is validated through simulations using channel models and 154~GHz corridor measurements.

We propose HuPER, a human-inspired framework that models phonetic perception as adaptive inference over acoustic-phonetics evidence and linguistic knowledge. With only 100 hours of training data, HuPER achieves state-of-the-art phonetic error rates on five English benchmarks and strong zero-shot transfer to 95 unseen languages. HuPER is also the first framework to enable adaptive, multi-path phonetic perception under diverse acoustic conditions. All training data, models, and code are open-sourced. Code and demo avaliable at https://github.com/Berkeley-Speech-Group/HuPER.

Rigorous uncertainty quantification is essential for the safe deployment of autonomous systems in unconstrained environments. Conformal Prediction (CP) provides a distribution-free framework for this task, yet its standard formulations rely on exchangeability assumptions that are violated by the distribution shifts inherent in real-world robotics. Existing online CP methods maintain target coverage by adaptively scaling the conformal threshold, but typically employ a static nonconformity score function. We show that this fixed geometry leads to highly conservative, volume-inefficient prediction regions when environments undergo structural shifts. To address this, we propose $\textbf{AdaptNC}$, a framework for the joint online adaptation of both the nonconformity score parameters and the conformal threshold. AdaptNC leverages an adaptive reweighting scheme to optimize score functions, and introduces a replay buffer mechanism to mitigate the coverage instability that occurs during score transitions. We evaluate AdaptNC on diverse robotic benchmarks involving multi-agent policy changes, environmental changes and sensor degradation. Our results demonstrate that AdaptNC significantly reduces prediction region volume compared to state-of-the-art threshold-only baselines while maintaining target coverage levels.

Camera-based visible light positioning (VLP) is a promising technique for accurate and low-cost indoor camera pose estimation (CPE). To reduce the number of required light-emitting diodes (LEDs), advanced methods commonly exploit LED shape features for positioning. Although interesting, they are typically restricted to a single LED geometry, leading to failure in heterogeneous LED-shape scenarios. To address this challenge, this paper investigates Lamé curves as a unified representation of common LED shapes and proposes a generic VLP algorithm using Lamé curve-shaped LEDs, termed LC-VLP. In the considered system, multiple ceiling-mounted Lamé curve-shaped LEDs periodically broadcast their curve parameters via visible light communication, which are captured by a camera-equipped receiver. Based on the received LED images and curve parameters, the receiver can estimate the camera pose using LC-VLP. Specifically, an LED database is constructed offline to store the curve parameters, while online positioning is formulated as a nonlinear least-squares problem and solved iteratively. To provide a reliable initialization, a correspondence-free perspective-n-points (FreePnP) algorithm is further developed, enabling approximate CPE without any pre-calibrated reference points. The performance of LC-VLP is verified by both simulations and experiments. Simulations show that LC-VLP outperforms state-of-the-art methods in both circular- and rectangular-LED scenarios. Compared to a perspective arcs algorithm, LC-VLP can achieve reductions of both over 30% in average position and rotation errors. Experiments further show that LC-VLP can achieve an average position accuracy of less than 4 cm.

Capturing display screens with mobile devices has become increasingly common, yet the resulting images often suffer from severe degradations caused by the coexistence of moiré patterns and flicker-banding, leading to significant visual quality degradation. Due to the strong coupling of these two artifacts in real imaging processes, existing methods designed for single degradations fail to generalize to such compound scenarios. In this paper, we present the first systematic study on joint removal of moiré patterns and flicker-banding in screen-captured images, and propose a unified restoration framework, named CLEAR. To support this task, we construct a large-scale dataset containing both moiré patterns and flicker-banding, and introduce an ISP-based flicker simulation pipeline to stabilize model training and expand the degradation distribution. Furthermore, we design a frequency-domain decomposition and re-composition module together with a trajectory alignment loss to enhance the modeling of compound artifacts. Extensive experiments demonstrate that the proposed method consistently. outperforms existing image restoration approaches across multiple evaluation metrics, validating its effectiveness in complex real-world scenarios.

The emergence of Large Audio-Language Models (LALMs) has advanced Speech Emotion Recognition (SER), but their size limits deployment in resource-constrained environments. While Knowledge Distillation is effective for LALM compression, existing methods remain underexplored in distilling the cross-modal projection module (Projector), and often struggle with alignment due to differences in feature dimensions. We propose PL-Distill, a KD framework that combines Projector-Level Distillation (PDist) to align audio embeddings and Logits-Level Distillation (LDist) to align output logits. PDist introduces Attention-weighted Centered Kernel Alignment, a novel approach we propose to highlight important time steps and address dimension mismatches. Meanwhile, LDist minimizes the Kullback-Leibler divergence between teacher and student logits from audio and text modalities. On IEMOCAP, RAVDESS, and SAVEE, PL-Distill compresses an 8.4B-parameter teacher to a compact 1.1B-parameter student, consistently outperforming the teacher, state-of-the-art pretrained models, and other KD baselines across all metrics.

This paper presents an LMI-based design framework for multirate steady-state Kalman filters in systems with sensors operating at different sampling rates. The multirate system is formulated as a periodic time-varying system, where the Kalman gains converge to periodic steady-state values that repeat every frame period. Cyclic reformulation transforms this into a time-invariant problem; however, the resulting measurement noise covariance becomes semidefinite rather than positive definite, preventing direct application of standard Riccati equation methods. I address this through a dual LQR formulation with LMI optimization that naturally handles semidefinite covariances. The framework enables multi-objective design, supporting pole placement for guaranteed convergence rates and $l_2$-induced norm constraints for balancing average and worst-case performance. Numerical validation using an automotive navigation system with GPS and wheel speed sensors, including Monte Carlo simulation with 500 independent noise realizations, demonstrates that the proposed filter achieves a position RMSE well below the GPS noise level through effective multirate sensor fusion, and that the LMI solution provides valid upper bounds on the estimation error covariance.

This paper proposes a novel hybrid control framework for switched linear parameter-varying (LPV) systems under hysteresis switching logic. By introducing a controller state-reset mechanism, the hybrid LPV synthesis problem is reformulated as a convex optimization problem expressed in terms of linear matrix inequalities (LMIs), enabling efficient computation of both switching LPV controller gains and reset matrices. The proposed approach is then applied to active magnetic bearing (AMB) systems, whose rotor dynamics exhibit strong dependence on rotational speed. Conventional LPV designs are often conservative due to large speed variations. The proposed hybrid gain-scheduled controller explicitly accounts for bounds on parameter variation rates, employs multiple LPV controllers over distinct operating regions, and uses hysteresis switching to reduce chattering and ensure stability. The effectiveness of the approach is demonstrated through a detailed AMB control design example.

We present a white-box adaptive NMPC architecture that resolves vehicular plasticity (adaptation to varying operating regimes without retraining) by arbitrating among frozen, regime-specific neural specialists using a Modular Sovereignty paradigm. The ensemble dynamics are maintained as a fully traversable symbolic graph in CasADi, enabling maximal runtime auditability. Synchronous simulation validates rapid adaptation (~7.3 ms) and near-ideal tracking fidelity under compound regime shifts (friction, mass, drag) where non-adaptive baselines fail. Empirical benchmarking quantifies the transparency cost: symbolic graph maintenance increases solver latency by 72-102X versus compiled parametric physics models, establishing the efficiency price of strict white-box implementation.

Semantic watermarks exhibit strong robustness against conventional image-space attacks. In this work, we show that such robustness does not survive under micro-geometric perturbations: spatial displacements can remove watermarks by breaking the phase alignment. Motivated by this observation, we introduce MarkCleaner, a watermark removal framework that avoids semantic drift caused by regeneration-based watermark removal. Specifically, MarkCleaner is trained with micro-geometry-perturbed supervision, which encourages the model to separate semantic content from strict spatial alignment and enables robust reconstruction under subtle geometric displacements. The framework adopts a mask-guided encoder that learns explicit spatial representations and a 2D Gaussian Splatting-based decoder that explicitly parameterizes geometric perturbations while preserving semantic content. Extensive experiments demonstrate that MarkCleaner achieves superior performance in both watermark removal effectiveness and visual fidelity, while enabling efficient real-time inference. Our code will be made available upon acceptance.

Data centers (DCs) are increasingly recognized as flexible loads that can support grid frequency regulation. Yet, most existing methods treat workload scheduling and regulation capacity bidding separately, overlooking how queueing dynamics and spatial-temporal dispatch decisions affect the ability to sustain real-time regulation. As a result, the committed regulation may become infeasible or short-lived. To address this issue, we propose a unified day-ahead co-optimization framework that jointly decides workload distribution across geographically distributed DCs and regulation capacity commitments. We construct a space-time network model to capture workload migration costs, latency requirements, and heterogeneous resource limits. To ensure that the committed regulation remains deliverable, we introduce chance constraints on instantaneous power flexibility based on interactive load forecasts, and apply Value-at-Risk queue-state constraints to maintain sustainable response under cumulative regulation signals. Case studies on a modified IEEE 68-bus system using real data center traces show that the proposed framework lowers system operating costs, enables more viable regulation capacity, and achieves better revenue-risk trade-offs compared to strategies that optimize scheduling and regulation independently.

Electrification of heavy-duty vehicles places substantial stress on distribution grids, and Charging Energy Hubs (CEHs) mitigate these impacts by integrating charging infrastructure with renewable energy sources and battery storage. Optimal sizing of CEH components is therefore a critical investment decision, yet challenging because design choices depend strongly on operational dynamics. This work presents a mixed-integer linear programming model for the optimal sizing of CEH components, using a co-design approach that jointly optimizes component sizing and operational decisions. A case study for a heavy-duty fleet demonstrates the effectiveness of the method for cost-efficient, scalable, and grid-compliant CEH planning.

In two-strategy decision-making problems, individuals often imitate the highest earners or choose either the common or rare strategy. Individuals who benefit from the common strategy are conformists, whereas those who profit by choosing the less common one are called nonconformists. The population proportions of the two strategies may undergo perpetual fluctuations in finite, discrete, heterogeneous populations of imitators, conformists, and nonconformists. How these fluctuations evolve as population size increases was left as an open question and is addressed in this paper. We show that the family of Markov chains describing the discrete population dynamics forms a generalized stochastic approximation process for a differential inclusion–the continuous-time dynamics. Furthermore, we prove that the continuous-time dynamics always equilibrate. Then, by leveraging results from the stochastic approximation theory, we show that the amplitudes of fluctuations in the proportions of the two strategies in the population approach zero with probability one when the population size grows to infinity. Our results suggest that large-scale perpetual fluctuations are unlikely in large, well-mixed populations consisting of these three types, particularly when imitators follow the highest earners.

This study proposes a novel adaptive traffic signal control method leveraging a Deep Q-Network (DQN) and Proximal Policy Optimization (PPO) to optimize signal timing by integrating variable cell length and multi-channel state representation. A road partition formula consisting of the sum of logarithmic and linear functions was proposed. The state variables are a vector composed of three channels: the number of vehicles, the average speed, and space occupancy. The set of available signal phases constitutes the action space, the selected phase is executed with a fixed green time. The reward function is formulated using the absolute values of key traffic state metrics - waiting time, speed, and fuel consumption. Each metric is normalized by a typical maximum value and assigned a weight that reflects its priority and optimization direction. The simulation results, using Sumo-TensorFlow-Python, demonstrate a cross-range transferability evaluation and show that the proposed variable cell length and multi-channel state representation method excels compared to fixed cell length in optimization performance.

Bird strikes pose a significant threat to aviation safety, often resulting in loss of life, severe aircraft damage, and substantial financial costs. Existing bird strike prevention strategies primarily rely on avian radar systems that detect and track birds in real time. A major limitation of these systems is their inability to identify bird species, an essential factor, as different species exhibit distinct flight behaviors, and altitudinal preference. To address this challenge, we propose an image-based bird classification framework using Convolutional Neural Networks (CNNs), designed to work with camera systems for autonomous visual detection. The CNN is designed to identify bird species and provide critical input to species-specific predictive models for accurate flight path prediction. In addition to species identification, we implemented dedicated CNN classifiers to estimate flock formation type and flock size. These characteristics provide valuable supplementary information for aviation safety. Specifically, flock type and size offer insights into collective flight behavior, and trajectory dispersion . Flock size directly relates to the potential impact severity, as the overall damage risk increases with the combined kinetic energy of multiple birds.

Emotion recognition from human speech is a critical enabler for socially aware conversational AI. However, while most prior work frames emotion recognition as a categorical classification problem, real-world affective states are often ambiguous, overlapping, and context-dependent, posing significant challenges for both annotation and automatic modeling. Recent large-scale audio language models (ALMs) offer new opportunities for nuanced affective reasoning without explicit emotion supervision, but their capacity to handle ambiguous emotions remains underexplored. At the same time, advances in inference-time techniques such as test-time scaling (TTS) have shown promise for improving generalization and adaptability in hard NLP tasks, but their relevance to affective computing is still largely unknown. In this work, we introduce the first benchmark for ambiguous emotion recognition in speech with ALMs under test-time scaling. Our evaluation systematically compares eight state-of-the-art ALMs and five TTS strategies across three prominent speech emotion datasets. We further provide an in-depth analysis of the interaction between model capacity, TTS, and affective ambiguity, offering new insights into the computational and representational challenges of ambiguous emotion understanding. Our benchmark establishes a foundation for developing more robust, context-aware, and emotionally intelligent speech-based AI systems, and highlights key future directions for bridging the gap between model assumptions and the complexity of real-world human emotion.

This paper proposes a unified family of learnable Koopman operator parameterizations that integrate linear dynamical systems theory with modern deep learning forecasting architectures. We introduce four learnable Koopman variants-scalar-gated, per-mode gated, MLP-shaped spectral mapping, and low-rank Koopman operators which generalize and interpolate between strictly stable Koopman operators and unconstrained linear latent dynamics. Our formulation enables explicit control over the spectrum, stability, and rank of the linear transition operator while retaining compatibility with expressive nonlinear backbones such as Patchtst, Autoformer, and Informer. We evaluate the proposed operators in a large-scale benchmark that also includes LSTM, DLinear, and simple diagonal State-Space Models (SSMs), as well as lightweight transformer variants. Experiments across multiple horizons and patch lengths show that learnable Koopman models provide a favorable bias-variance trade-off, improved conditioning, and more interpretable latent dynamics. We provide a full spectral analysis, including eigenvalue trajectories, stability envelopes, and learned spectral distributions. Our results demonstrate that learnable Koopman operators are effective, stable, and theoretically principled components for deep forecasting.

Identifying the dynamic precompensator that renders a nonlinear control system feedback linearizable is a challenging problem. Researchers have explored the problem – dynamic feedback linearization – and produced existence conditions and constructive procedures for the dynamic precompensator. These remain, in general, either computationally expensive or restrictive. Treating the challenge as intrinsic, this article views the problem as a search problem over a category. Dynamic programming applies and, upon restriction to a finite category, classic search algorithms find the minimal dynamic extension. Alternatively, a heuristic aiming towards feedback linearizable systems can be employed to select amongst the infinitely-many extensions. This framing provides a distinctive, birds-eye view of the search for the dynamic precompensator.

Ultrasound is the most widely used medical imaging modality, yet the images it produces are fundamentally unique, arising from tissue-dependent scattering, reflection, and speed-of-sound variations that produce a constrained set of characteristic textures that differ markedly from natural-image statistics. These acoustically driven patterns make ultrasound challenging for algorithms originally designed for natural images. To bridge this gap, the field has increasingly turned to foundation models, hoping to leverage their generalization capabilities. However, these models often falter in ultrasound applications because they are not designed for ultrasound physics, they are merely trained on ultrasound data. Therefore, it is essential to integrate ultrasound-specific domain knowledge into established learning frameworks. We achieve this by reformulating self-supervised learning as a texture-analysis problem, introducing texture ultrasound semantic analysis (TUSA). Using TUSA, models learn to leverage highly scalable contrastive methods to extract true domain-specific representations directly from simple B-mode images. We train a TUSA model on a combination of open-source, simulated, and in vivo data. The latent space is compared to several larger foundation models, demonstrating that our approach gives TUSA models better generalizability for difficult downstream tasks on unique online datasets as well as a clinical eye dataset collected for this study. Our model achieves higher accuracy in detecting COVID (70%), spinal hematoma (100%) and vitreous hemorrhage (97%) and correlates more closely with quantitative parameters like liver steatosis (r = 0.83), ejection fraction (r = 0.63), and oxygen saturation (r = 0.38). We open-source the model weights and training script: https://github.com/talg2324/tusa

This paper studies preview control in both the $H_\infty$ and regret-optimal settings. The plant is modeled as a discrete-time, linear time-invariant system subject to external disturbances. The performance baseline is the optimal non-causal controller that has full knowledge of the disturbance sequence. We first review the construction of the $H_\infty$ preview controller with $p$-steps of disturbance preview. We then show that the closed-loop $H_\infty$ performance of this preview controller converges as $p\to \infty$ to the performance of the optimal non-causal controller. Furthermore, we prove that the optimal regret of the preview controller converges to zero. These results demonstrate that increasing preview length allows controllers to asymptotically achieve non-causal performance in both the $H_\infty$ and regret frameworks. A numerical example illustrates the theoretical results.

This paper addresses the challenge of audio-visual single-microphone speech separation and enhancement in the presence of real-world environmental noise. Our approach is based on generative inverse sampling, where we model clean speech and ambient noise with dedicated diffusion priors and jointly leverage them to recover all underlying sources. To achieve this, reformulate a recent inverse sampler to match our setting. We evaluate on mixtures of 1, 2, and 3 speakers with noise and show that, despite being entirely unsupervised, our method consistently outperforms leading supervised baselines in WER across all conditions. We further extend our framework to handle off-screen speaker separation. Moreover, the high fidelity of the separated noise component makes it suitable for downstream detection of the acoustic scene. Code and pretrained models will become available upon acceptance. Demo page: https://ssnaps2026.github.io/ssnaps2026/

Gaussian Mixture Models (GMMs) commonly arise in communication systems, particularly in bilinear joint estimation and detection problems. Although the product of GMMs is still a GMM, as the number of factors increases, the number of components in the resulting product GMM grows exponentially. To obtain a tractable approximation for a univariate factored probability density function (PDF), such as a product of GMMs, we investigate iterative message-passing algorithms. Based on Belief Propagation (BP), we propose a Variable Duplication and Gaussian Belief Propagation (VDBP)-based algorithm. The key idea of VDBP is to construct a multivariate measurement model whose marginal posterior is equal to the given univariate factored PDF. We then apply Gaussian BP (GaBP) to transform the global inference problem into local ones. Expectation propagation (EP) is another branch of message passing algorithms. In addition to converting the global approximation problem into local ones, it features a projection operation that ensures the intermediate functions (messages) belong to a desired family. Due to this projection, EP can be used to approximate the factored PDF directly. However, even if every factor is integrable, the division operation in EP may still cause the algorithm to fail when the mean and variance of a non-integrable belief are required. Therefore, this paper proposes two methods that combine EP with our previously proposed techniques for handling non-integrable beliefs to approximate univariate factored distributions.

Self-supervised speaker embeddings are widely used in speaker verification systems, but prior work has shown that they often encode sensitive demographic attributes, raising fairness and privacy concerns. This paper investigates the extent to which demographic information, specifically gender, age, and accent, is present in SimCLR-trained speaker embeddings and whether such leakage can be mitigated without severely degrading speaker verification performance. We study two debiasing strategies: adversarial training through gradient reversal and a causal bottleneck architecture that explicitly separates demographic and residual information. Demographic leakage is quantified using both linear and nonlinear probing classifiers, while speaker verification performance is evaluated using ROC-AUC and EER. Our results show that gender information is strongly and linearly encoded in baseline embeddings, whereas age and accent are weaker and primarily nonlinearly represented. Adversarial debiasing reduces gender leakage but has limited effect on age and accent and introduces a clear trade-off with verification accuracy. The causal bottleneck further suppresses demographic information, particularly in the residual representation, but incurs substantial performance degradation. These findings highlight fundamental limitations in mitigating demographic leakage in self-supervised speaker embeddings and clarify the trade-offs inherent in current debiasing approaches.

In this technical communique, we generalize the well-known Lyapunov-based stabilizability and detectability tests for discrete-time linear time-invariant systems to polytopic linear parameter-varying systems using the class of so-called poly-quadratic Lyapunov functions.