Papers

Automatic Modulation Recognition (AMR) is a key enabling technology for cognitive radio and intelligent spectrum management in next-generation wireless systems. However, current deep learning-based AMR methods predominantly rely on static multi-scale fusion strategies, which lack the flexibility to adapt to the highly dynamic temporal variations of modulation signals. To address this limitation, we propose MoEformer, an adaptive Multi-Scale Mixture-of-Experts Transformer network that directly processes I/Q signals to preserve their temporal and phase structures. Specifically, MoEformer constructs multi scale expert views through temporal resampling, employs an input-dependent gating mechanism for dynamic expert fusion, and integrates Rotary Position Embeddings (RoPE) within Transformer encoders to capture both local and global tem poral dependencies. Comprehensive evaluations on three widely adopted benchmarks (RadioML2016.10a, RadioML2016.10b, and RadioML2018.01A) demonstrate that MoEformer outperforms the competitive baselines, achieving superior average recognition accuracies of 63.74%, 66.24%, and 64.22%, respectively. In addition, the proposed method strikes an optimal trade-off between recognition performance and model complexity.

Streaming zero-shot voice conversion (VC) has become increasingly popular due to its potential for real-time applications. The recently proposed MeanVC achieves lightweight streaming zero-shot VC, but it has several limitations: its chunk-wise autoregressive denoising doubles the effective training sequence length, conversion quality degrades under small-chunk settings, and its timbre encoder directly relies on reference mel-spectrograms, making it sensitive to reference audio quality. To address these limitations we propose MeanVC 2. We introduce future-receptive chunking (FRC), which explicitly schedules past and future receptive fields across diffusion transformer decoder layers and removes clean-chunk teacher forcing. By incorporating bounded future context, FRC enables stable conversion with a 40 ms chunk size. We further introduce a universal timbre token encoder, which constructs a timbre representation from a global speaker embedding and retrieves fine-grained timbre cues via cross-attention, improving robustness to low-quality references and enhancing zero-shot speaker similarity. Experimental results show that MeanVC 2 significantly outperforms MeanVC, while reducing latency from 211 ms to 110 ms. Audio samples are publicly available. The source code will be publicly released.

Removing intermediate representations and separately trained decoding stages has become an important direction in generative modeling. In text-to-speech, however, high-quality systems are still commonly built through an intermediate acoustic representation before waveform synthesis. In this work, we present BareWave, a fully waveform-native framework for direct text-to-wave generation in flow-matching TTS. We consider this setting to raise three training challenges: raw-waveform modeling lacks a strong pretrained representational scaffold, different stages of training benefit from different noise schedules, and data-space perceptual objectives do not automatically share the temporal structure of the velocity-space flow objective. As a result, direct waveform training is hard to optimize efficiently, hard to push toward a strong final operating point with a fixed recipe, and hard to integrate effective perceptual refinement. Guided by this view, we develop a direct text-to-wave training framework that combines training-time representation alignment, staged noise scheduling, and velocity-aware perceptual alignment (VAPA), while preserving a single waveform-native inference path without pretrained components at test time. Experiments on zero-shot voice cloning show that strong intelligibility, speaker similarity, and naturalness can be achieved under a fully waveform-native inference path, supporting waveform-native flow-matching TTS as a practical direction. Project page with audio demos is available at https://barewave.github.io/.

A classical universal stabilization formula offers the practitioner no design freedom: it is a single, parameter-free object. We introduce a cost-parametrized family of stabilizing feedback laws, where (1) the user chooses a function that serves as the running cost on control in an inverse-optimal cost functional, and (2) obtains, through a formula, a nonlinear "expander" of a pre-existing universal controller, which solves an infinite-horizon optimal control problem with a meaningful cost on the state. The cost-to-expander formula is a three-step construction, involving, inter alia, cost differentiation and function inversion-overall, a nonlinear infinite-dimensional operator. The cost-to-expander operator is proven Lipschitz, which enables uniform neural operator approximation of the entire family and supports both offline performance exploration and online adaptation. Semiglobal practical asymptotic stability and second-order suboptimality bounds are established under the approximation. The operator learning and its use in semiglobal stabilization are illustrated numerically. We call the result 'half-direct-optimal' because the paper's design is less than a general 'direct optimal' (HJB-inducing) control, but more than the fully inverse optimal, since the user performs minimization for an arbitrary given cost on control. The dual to the half-direct problem we solve is the problem in which the cost on the state is arbitrary and given. This dual problem is easier and outside of the scope of the paper.

Unplanned grid-connected (GC)/stand-alone (SA) transitions commonly occur in AC microgrids during protection trips, manual breaker operation, or low-bandwidth supervisory communication. Under such unplanned transitions, a grid-forming inverter must support the local-load voltage in stand-alone operation and regulate the desired power/current injection in grid-connected operation. Existing P–Q droop-based seamless-transfer methods often rely on planned transition commands, supervisory islanding detection, or pre-synchronization interval, which may prevent timely voltage/current support during unplanned bidirectional transitions. To address this problem, this paper proposes a seamless contraction-control (SCC) framework for target dynamics. Using the SCC, contraction-based grid-connected current-control and stand-alone voltage-control laws are proposed. With the new control laws, the inverter achieves transient stability and converges to the target trajectory with a prescribed convergence rate. Furthermore, a breaker-status observer is proposed to infer the grid-connected/stand-alone mode from voltage measurements on both sides of the breaker, eliminating the need for a dedicated pre-synchronization interval or supervisory islanding detection process and enabling timely voltage/current support during unplanned transitions. Experimental results validate that the proposed method achieves stand-alone voltage support, stable grid-connected current injection under symmetrical/unsymmetrical grid-voltage sag and phase-jump disturbances, and unplanned bidirectional transitions.

We propose LEAF, a learning-enabled ADMM framework for accelerated convex optimization. The key idea is to approximate the Moreau envelope of the objective function using an Input Convex Neural Network (ICNN), resulting in a learned model that preserves convexity and smoothness. This leads to the proposed Moreau Envelope Learning ADMM (MEL-ADMM) and its splitting variant sMEL-ADMM. Unlike existing approaches that learn high-dimensional operators directly, LEAF learns a scalar-valued Moreau envelope, significantly reducing model complexity and improving data efficiency. The framework accommodates a broad class of convex problems with smooth and non-smooth objectives. By embedding convexity explicitly through the ICNN architecture, the proposed approach maintains high approximation accuracy while preserving key structural properties of the optimization problem. Both MEL-ADMM and sMEL-ADMM are developed with theoretical guarantees of convergence and feasibility under the learned model. Rigorous analysis shows that the proposed methods achieve convergence rates comparable to classical ADMM while reducing per-iteration computational cost. Numerical experiments demonstrate up to an order-of-magnitude speedup over state-of-the-art solvers while maintaining low optimality gaps

Warm starts are widely used to accelerate AC optimal power flow (ACOPF) solves, but the impact of different initialization strategies has received limited systematic study, particularly for the primal-dual interior-point methods that dominate large-scale ACOPF algorithms. This paper benchmarks initialization strategies for ACOPF solved with the interior-point solver IPOPT on 19 PGLib-OPF instances (5 to 30,000 buses), testing all 15 non-empty subsets of the primal blocks $\{P_g, Q_g, V_m, V_a\}$ under oracle conditions and three DC-seeded combinations in a practical setting. The experiments show that most partial primal-plus-dual restarts increase solve time or reduce convergence reliability. Among the oracle primal-plus-dual (O-PD) configurations, only the complete restart reliably converges on every baseline-convergent case, reaching a $47.6\%$ median solve-time speedup. Twelve of the 14 partial O-PD combinations have negative median speedups, and several fail repeatedly on larger networks. Decomposing the dual into constraint and bound multipliers shows that coverage, not the presence of duals per se, governs robustness: the full bound-multiplier vector reaches 90.7% convergence and a $+26.8$% median speedup, whereas block-matched coverage (oracle multipliers on some bounds, defaults on the rest) drops to 70.4% and $-31.1$%. Practical DC seeding sometimes helps the AC solve, but the benefit is no longer statistically significant once the DCOPF presolve cost is included in the end-to-end comparison ($p = 0.4171$). For learned warm-start methods, the results support the following target ordering: predict the full primal vector first; if only partial coverage is possible, prioritize voltage variables; and avoid partial or inconsistent dual predictions unless the primal estimate is nearly complete.

In the task of few-shot class-incremental audio classification, the number of classes is assumed to always increase without considering the possibility of decrease. However, the number of classes generally increases or decreases in practice. In this paper, we investigate a problem of Few-shot Class-variable Incremental Audio Classification (FCIAC), in which the number of classes increases or decreases. We propose a FCIAC method using prototype adaptation and pseudo class-variable training. The model in our method consists of an encoder and a classifier. The classifier is initialized by a class-variable prototype adaptation network, whose structure dynamically changes with the change of classes. In addition, we design a pseudo class-variable training strategy to enhance the model's adaptability to changing classes. Experiments on three public datasets show that our method exceeds previous methods in average accuracy. The code is at: https://github.com/cgq2971-afk/FCIAC.

Conventionally, existing vehicle platooning approaches are designed for connected vehicles, typically including connected autonomous vehicles and connected human-driven vehicles. Non-connected vehicles, such as non-connected autonomous or human-driven vehicles, are not incorporated. As a result, these platooning approaches may not properly reflect real-world mixed traffic conditions at the current stage. To address this limitation, this study proposes a hybrid platooning pattern that conditionally permits non-connected vehicles to join platoons, thereby enhancing platooning diversity and flexibility. However, it was found that the unregulated integration of non-connected vehicles can trigger rapid platoon expansion, significantly amplifying the risk of disturbance propagation in traffic flow. This, in turn, exacerbates the inherent conflict between traffic throughput and stability. To mitigate these challenges, this paper further develops a hybrid platooning control strategy based on deep reinforcement learning (DRL). This strategy integrates vehicle dynamics, platoon topology, and traffic flow states through a multi-level state representation network, enabling a dynamic trade-off between traffic capacity and stability. Numerical simulations demonstrate that the proposed strategy effectively suppresses velocity disturbance propagation by dynamically optimizing platoon structures, thereby significantly enhancing the stability and safety of mixed traffic while reducing fuel consumption and emissions.

The analysis of non stationary signals in complex physical systems often relies on Time Frequency distributions. Among these, the Pseudo Wigner Ville Distribution (PWVD) stands out for its superior resolution but is mathematically challenging due to its inherent quadratic nonlinearity. This nonlinearity generates complex interference artifacts and cross terms in the phase space, potentially obscuring the physical features of the signal, particularly for nonlinear chirps. In this work, we establish a mathematically grounded framework for the PWVD for general windowed nonlinear chirps. By leveraging the theory of oscillatory integrals with coalescing stationary points, we derive a uniform asymptotic expansion that bridges the gap between heuristic signal processing and semiclassical geometric approaches (Berry's chord construction). The resulting closed form representation, expressed in terms of symmetric incomplete Airy functions, provides a unified description of the nonlinear transform's behavior, regularizing the transition across the instantaneous frequency caustics. While the framework is general, we show its power on two illustrative examples: the high precision nonlinear chirps of coalescing binaries in gravitational-wave astronomy and radar nonlinear chirps for pulse compression applications. The analytical results successfully predict the structure of interference patterns and quantify the systematic bias in peak based frequency estimation. Therefore, this study establishes a systematic bridge between nonlinear mathematical analysis and precision experimental physics, validating the PWVD as a robust tool for detailed source characterization in high signal to noise regimes.

Continuous, non-contact blood pressure (NCBP) monitoring holds significant promise for pervasive cardiovascular care, yet single-modality approaches – such as imaging photoplethysmography (iPPG) – remain constrained by environmental artifacts, skin-tone sensitivity, and the absence of proximal cardiac mechanical information. This study investigates the feasibility of a dual-modality sensing paradigm that synergistically integrates facial iPPG with posterior-facing frequency-modulated continuous wave (FMCW) millimeter-wave radar to capture complementary hemodynamic cues: distal optical volumetric fluctuations and proximal cardiac micro-motions (radar motion signals, RMS). To bridge the morphological disparity between these heterogeneous streams, we develop an end-to-end deep learning architecture, BiLSTM-MS-DiCNN, which leverages multi-scale dilated convolutions for spatial feature extraction and bidirectional long short-term memory for temporal dependency modeling. In a controlled feasibility study involving 15 healthy participants across distinct hemodynamic states (resting, deep breathing, and post-exercise), the proposed framework achieved a Mean Absolute Difference (MAD) of 4.71 mmHg for systolic BP (SBP) and 4.60 mmHg for diastolic BP (DBP) under resting conditions, with consistent performance during physiological perturbations. These preliminary findings demonstrate the viability of mmWave-iPPG fusion as a promising pathway toward robust, unobtrusive NCBP monitoring.

Improving energy efficiency in the transportation sector is critical for achieving sustainable mobility, with eco-driving emerging as a key strategy. However, implementing effective eco-driving for connected and automated vehicles (CAVs) in mixed traffic presents a significant control challenge due to the heterogeneous, uncertain behavior of human-driven vehicles (HDVs). Data-enabled Predictive Control (DeePC) offers a promising model-free approach but is often hindered by a high computational burden, limiting its real-time feasibility. This paper introduces a novel Direct Data-driven Predictive Control (D3PC) framework to address this limitation. By reformulating the data-driven prediction mechanism, the D3PC significantly reduces computational complexity, making its computation time nearly invariant to historical data size. This computational efficiency directly enables the formulation of a sophisticated eco-driving controller that can solve the complex energy optimization problem in real time, even within diverse and stochastic mixed-traffic environments. Comprehensive simulations demonstrate that the D3PC is orders of magnitude faster than existing DeePC-based methods while achieving superior energy efficiency. Specifically, it reduces total platoon energy consumption by up to 10.71% compared to rule-based cruise control baselines and 3.80% compared to the original DeePC, confirming its effectiveness for real-time, energy-efficient control.

We develop a unified framework for linear-cost optimal control, finite-time optimal steering, dissipativity analysis, and zero-sum differential games for linear impulsive systems whose state is a Hermitian matrix evolving in $\mathbb{H}^{n+m}_{\succeq0}$, a class that encompasses continuous- and discrete-time linear systems and switched systems as degenerate cases, and includes the second-order moment dynamics of linear (stochastic) hybrid systems. The entire theory rests on three tools: a single key identity relating cost, trajectory, and a dual variable, an Extended Schur complement lemma, and a Schur inner-product decomposition, applied identically to the flow integral and to each jump. These yield structurally uniform sufficient and necessary conditions, dual linear matrix inequality (LMI) characterizations, and explicit optimal policies for every problem class, on both finite and infinite horizons under time-varying assumptions (without time invariance or periodicity), together with causal dwell-time policies for the problems that admit them.

This paper presents an adaptive model predictive control (MPC) framework for nonlinear urban air mobility (UAM) vehicles operating across the full flight envelope. The proposed approach leverages a linear parameter-varying (LPV) representation to update the predictive model online, enabling accurate capture of strongly nonlinear and time-varying dynamics associated with distributed electric propulsion (DEP) eVTOL aircraft. To systematically address the high-dimensional and coupled nature of MPC tuning, a multi-objective evolutionary optimization strategy based on NSGA-II is employed, incorporating proper normalization of states and control inputs to ensure balanced weighting and meaningful exploration of the design space. The resulting controller explicitly accounts for actuator constraints and enables reconfigurable control allocation for fault-tolerant operation. The framework is evaluated in nonlinear simulations using NASA's Generic Urban Air Mobility (GUAM) model and benchmarked against a robust servomechanism linear quadratic regulator (RSLQR). Results demonstrate that the proposed adaptive MPC achieves improved trajectory tracking and enhanced robustness under both nominal conditions and actuator degradation scenarios, including partial motor failure, while maintaining constraint satisfaction throughout all flight regimes.

The energy transition in power systems requires flexible assets to offset renewable generation variability across multiple time scales, while supporting the integration of renewables and the electrification of demand without requiring costly grid reinforcement. Energy storage occupies a unique position among these assets: depending on the technology, it can provide short-duration grid services at high ramping rates, such as frequency regulation and voltage support, longer-duration functions such as intra-day peak shaving, or inter-seasonal energy buffering. This multi-service character, combined with the declining costs of energy storage technologies (most notably that of battery energy storage systems), is central to the economic viability of storage investments. The value of a given installation depends strongly on its grid connection point and intended use case: an asset-coupled battery serving a consumer or generation plant faces a different service landscape, and therefore a different business case, than a network-coupled system operating as an independent grid resource. This paper presents a structured taxonomy of grid-connected energy storage applications, discusses the principal application domains, and describes the key challenges that must be addressed to integrate storage effectively into power systems. Services are discussed with special emphasis on the Swiss regulatory context. Finally, the STORE flagship project supported by the Swiss Innovation Agency (Innosuisse), where some of the critical challenges of energy storage integration in power grids are addressed, is introduced.

This work develops several constructive foundations for systems and control within Bishop-style constructive mathematics. For an engineer, the guiding principle is that an object claimed to exist, such as a trajectory, an optimal control law, a selector, or a viable solution, should come with finite data and an operation computing approximations to any prescribed precision. The style remains close to classical analysis, but existential statements are organized so that their computational content is visible. The paper begins with elementary geometric data in finite-dimensional Euclidean spaces: blocks, multiblocks, representable sets, regular functions, and certified integrals. This set-first integration route is meant to complement, rather than replace, abstract constructive integration theories such as Daniell-type or integration-space approaches. The developed apparatus is then applied to a constructive functional extremum-value theorem, selector extraction for multifunctions, Filippov-type and viable solutions of differential inclusions, regular probability densities, controlled Markov chains, and empirical density certificates. A short account of resolvent projectors and linear stability is included for completeness.

Deep learning (DL) methods are challenged to demonstrate robust performance across different segmentation datasets due to domain shifts, but active domain adaptation techniques enhance their generalization performance by querying a few samples from target domains for adaptation training. However in clinical practice, target domains often include private classes of new anatomical structures or pathologies that are not presented in the source data, and existing methods implement closed-set segmentation where source and target domains have the same segmentation classes. Additionally, source data are often inaccessible during adaptation due to strict data privacy regulations. To address these limitations, we propose an Active Source-free Open-set Domain Adaptation (ASFOSDA) method which is the first work to implement active learning for adapting DL models in open-set medical image segmentation without the access to source data. This method employs an active open-set query strategy to select the most informative target samples for training models based on Class-aware Decomposed Uncertainty (CDU) and Class-agnostic Prototype Discrepancy (CPD). CDU measures sample aleatoric uncertainty and model epistemic uncertainty by employing test time augmentation in stochastic processes. CPD measures cross-domain and self-domain discrepancy for selecting diverse samples. Subsequently, to boost the adaptation performance by enhancing training samples, a Target-refined Self-training strategy is proposed to generate high-quality pseudo labels for unselected samples, thus combining them with labeled samples for a semi-supervised training. We evaluated our method on cross-domain open-set volumetric medical image segmentation tasks, and it outperformed state-of-the-art adaptation methods.

In teleoperation, the human operator typically controls only the end-effector pose, which often leads to self-collisions of the manipulator and collisions with environmental obstacles, since joints and links are not controlled individually. A common strategy to mitigate this issue is to enhance the operator's input using optimal-control-based trajectory planning. As derivative-based solvers require differentiable constraints, existing approaches either approximate robots and obstacles with spheres, reducing geometric accuracy, or approximate derivatives, degrading convergence and increasing computation times. We address these limitations by adapting a recent formulation of differentiable collision-avoidance constraints, based on duality in convex optimization, to the teleoperation setting. The robot is approximated with capsules and the environment with polytopes. We compare the resulting trajectory planning method against state-of-the-art techniques in simulation with varying numbers of obstacles and evaluate it on a UR5e manipulator in a real-world teleoperation test. Results show that our approach achieves lower computation times while enabling more accurate obstacle modeling, leading to smoother and collision-free end-effector teleoperation.

Multirotors are widely used in applications ranging from surveillance to precision agriculture, yet conventional designs remain limited by their under-actuation. Tilt-rotor configurations overcome this limitation by enabling full actuation. This paper investigates neural-network-based control strategies for a fully actuated tilt-rotor system with four thrust-vectoring inputs. Our work is structured in two parts. First, we deliberately present a negative result by evaluating a direct input-output control approach. In this method, multilayer perceptrons (MLPs), long short-term memory (LSTM) networks, and transformer models are trained to map system states and their desired values directly to control signals. We show that this strategy fails to stabilize the system, highlighting the inherent difficulty of applying direct input-output learning to highly unstable plants. Second, as the main contribution, we propose a neural-network-enhanced sliding mode controller (SMC). The method decomposes the system dynamics into input-independent and input-dependent components, with the former learned from a small dataset using lightweight networks, thereby reducing real-time computational demands. Moreover, the proposed method can be trained using flight logs collected from low-performance controllers, and the resulting dynamic model learned from real-world data can be used in simulation. We further compare MLP- and LSTM-based implementations under model uncertainties and external disturbances, demonstrating the robustness and effectiveness of the proposed approach; in particular, the controller with the LSTM plant dynamics predictor achieves superior performance to its MLP-based counterpart while also exhibiting lower runtime.

AI-generated music detectors can appear robust on standard benchmark splits, yet their deployments require transfer to generator sources absent during training. We study this problem with source-restricted evaluation on MoM-open, an open reconstruction of MoM-CLAM that replaces the non-redistributable real corpus with FMA and MTG-Jamendo while preserving the fake-generator protocol. To isolate the role of representation, we introduce CoMoE, a compact fixed classifier for comparing heterogeneous audio token spaces while keeping the downstream architecture and training recipe unchanged. Experiments show that standard and real-source-restricted splits are nearly saturated, whereas fake-source restriction exposes large differences between token spaces: X-Codec tokens are strongest when training on Udio alone, while MERT-derived tokens are stronger when training on Suno-v3.5 alone. These results suggest that codec-style discrete token spaces should be treated as a primary experimental axis under generator shift in AI-generated music detection. Our code and data are available at https://github.com/MAAP-LAB/CoMoE.

We look at Bell inequalities from the lens of information structures in stochastic teams. We consider the usual Clauser-Horne-Shimony-Holt (CHSH) game and a dynamic variant of the same to study how various classes of strategies, classical, projective and quantum, behave under team theoretic solution concepts. We endow players with a fixed quantum state that we take as the Bell state, and allow variation of the measurement strategies. In the static game, projective strategies (where each player performs projective measurements) are team optimal for all parameter values and are closed under best response. In the dynamic game these properties fail: projective strategies may be strictly outperformed by classical ones, best responses to projective strategies need not be projective, and, beyond a sharp threshold on the channel noise, team optimal projective strategies cease to be person-by-person optimal, even though person-by-person optimal projective pairs always exist. These results shed light on the delicate interplay of information structure in quantum strategies and the fragility, under a change of information structure, of structural properties that are taken for granted in the static setting.

Protecting an asset against threats is a challenging problem in an era of continuously evolving intelligent attacks. This requires cooperation between the asset and the defender to share information and jointly maneuver. To address this problem, this work proposes a cooperative guidance and control strategy for active asset protection against a maneuvering threat. This work develops a joint maneuver strategy where both the defender and the asset coordinate their time-varying speeds and courses to neutralize/capture the attacker. The control strategy is formulated around three coupled geometric and temporal objectives. The first objective is to set the line-of-sight rate between the asset and the attacker to zero, putting the attacker on a collision course and reducing their maneuvering. The second objective is to maintain the defender on the line-of-sight between the asset and the attacker. This ensures that the attacker faces the defender first before reaching the vicinity of the asset. Lastly, the defender is also guided to pursue the attacker based on the time-to-go estimates between the defender and the attacker. While keeping these objectives in mind, the control actions for the asset and the defender are jointly designed, fostering cooperation between the two. The stability of the proposed strategy is established using a Lyapunov-based approach. Numerical simulations performed show the effectiveness of the proposed cooperative strategy in ensuring the successful capture of a maneuvering threat.

Whereas coil positioning in transcranial magnetic stimulation (TMS) to reach a specific cortical target with modern focal stimulation coils has been intensively studied, the alignment and contact of a coil with the head is often ignored. Focal figure-of-eight coils have a point on the surface, where they generate the largest induced electric field. This point should touch the head first, and the coil should be approximately tangential to the head in this point. Previous research has demonstrated the large impact if the coil does not touch the head with the right point and that many operators struggle with establishing or maintaining the correct coil-scalp alignment. This paper presents a technological support technology that can monitor the exact position of the contact point and also pressure to provide feedback to users. As the system uses exclusively components from consumer electronics, the sensor is low-cost and affordable. Through proper design, we achieved sufficient robustness so that the sensor does neither reset during TMS pulses and also not show any detectable degradation.

Cloning refers to producing identical copies of existing objects. Here, we experimentally show how to clone acoustic scattering objects. We acquire a digital twin and bring it back to life - a simple two-step process. First, we use broadband speakers to illuminate the scattering object within a closed receiver aperture. From these recorded reverberative data, we retrieve the object's scattering Green's functions using multidimensional deconvolution. In the second step, the acoustic scatterer is holographically reconstructed using the acquired scattering Green's functions. The hologram scatters any wavefield in real-time exactly like the original object would. Low-latency feedback reproduces all orders of interactions between the physical wavefield and the numerically defined hologram. This two-step process is demonstrated by cloning and modifying several rigid scatterers in a two-dimensional acoustic waveguide. Applications range from fully realistic digital scattering models to efficient metamaterial experimentation.

We study data-driven real-time economic optimization of a multi-product chemical reactor when no reliable first-principles model is available beyond a steady-state energy balance. Instead of learning the economic objective directly as a black-box function, we use a composite formulation in which Gaussian process (GP) models predict physically meaningful outputs, including product concentrations and reactor temperature, while profit is computed analytically from these predictions together with raw-material, product, and utility prices. This preserves the structure of the economic objective, makes it parametric in changing prices without needing retraining, and allows candidate operating points to be checked against the available energy balance through a physics residual. The GPs also provide predictive uncertainty, which is exploited in a Bayesian optimization (BO) framework both for data-efficient exploration and for conservative enforcement of the reactor temperature constraint through an upper confidence bound. The acquisition function additionally penalizes large energy-balance mismatch obtained by substituting the GP-predicted outputs and candidate inputs into the available steady-state energy balance. The approach is demonstrated on a benchmark simulation of a non-isothermal multi-product reactor. Relative to a trust-region safe BO implementation, the proposed method achieves better simulated economic performance within the available iteration budget. Relative to a purely data-driven BO approach that does not use the available physics information, it avoids reactor temperature constraint violations.

Deep learning EEG denoising architectures have scaled from tens of thousands to tens of millions of parameters, yet no prior study has isolated model capacity as the experimental variable or tested whether reconstruction metrics predict downstream neural-signal utility. We address both gaps by fixing architecture, loss, data split, and training recipe while sweeping only channel width from 1.05K to 40.26K parameters in a minimal depthwise-separable convolutional U-Net. Models were evaluated on the EEGDenoiseNet benchmark, cross-dataset BCI transfer tests, controlled baseline retraining, and downstream motor-imagery classification with five decoder families across all nine BCI Competition IV-2a subjects. Reconstruction performance saturated by 3-6.5K parameters, with post-elbow gains of at most 0.015 correlation coefficient per log10-parameter unit. An 8.46M-parameter baseline retrained under the same pipeline matched the 40.26K compact variant on EOG–a 200x parameter gap yielding no advantage–while a Patch-Transformer control reproduced the same diminishing-return shape. Downstream evaluation exposed a classifier-dependent metric-utility gap: reconstruction-optimized denoising significantly degraded CSP+LDA classification across all nine subjects and three artifact types (best denoised accuracy 0.547 vs. 0.612 noisy baseline; Bonferroni p=0.0488), persisting on naturally recorded trials (Delta=-0.047; BH-FDR q=0.0049). End-to-end neural decoders showed variable or neutral effects. Standard EEG denoising benchmarks are saturated far below current model capacity, and reconstruction metrics do not predict BCI utility. Ultra-compact models at 33-46 KB and 1.27-2.61M FLOPs/segment are practical for edge deployment. These findings argue for capacity-controlled evaluation, harder task-aware benchmarks, and mandatory downstream validation.

Deep learning on physiological time series is interpreted through domain-specific features – oscillatory rhythms in EEG, morphological complexes in ECG – yet these signals sit atop a broadband aperiodic 1/f-like envelope that covaries with arousal, age, and pathology. We introduce a spectral audit framework combining aperiodic/periodic decomposition, phase-preserving Fourier interventions, sham controls, and simulation validation. Aperiodic reliance was task-dependent and architecture-general: across six neural architectures, flattening drops exceeded 0.42 balanced-accuracy points for sleep-wake classification, reached 0.07-0.13 for clinical abnormality detection, and remained minimal for motor imagery. Six of seven EEG foundation models showed FDR-significant aperiodic reliance on clinical EEG; age/sex and recording-era controls reduced but did not eliminate the effect. Applying the audit to PTB-XL ECG revealed neural drops of 0.32–0.36 persisting after demographic matching, confirming this confound class extends beyond EEG. Aperiodic controls should become standard for interpretable physiological time-series deep learning.

Using speaker embeddings as conditioning can strengthen speech enhancement, but most methods either require clean enrollment audio or rely on embeddings extracted from noisy speech, which are fragile under noise and domain shift. We propose G-MaP-SE, a guided enhancement framework that builds a clean-speech embedding prior with a Gaussian Mixture Model (GMM) and refines a noisy conditioning embedding by matching it to this prior. The matched prior embedding is then injected into a time-frequency enhancement backbone via a lightweight gated fusion module. Experiments on VoiceBank+DEMAND and DNS Challenge 2020 datasets show that the proposed prior matching consistently outperforms noisy conditioning and substantially narrows the gap to an oracle clean-conditioning upper bound, while requiring no enrollment audio at inference time. The code, audio samples, and checkpoint are available.

Cloaking and holography – usually treated as distinct problems – are two limits of a single operation that we call acoustic disguising, realized here using immersive boundary conditions on a closed surface. Driving the boundary with homogeneous Green's functions suppresses any incident field inside the enclosed volume and cloaks unknown objects broadband; driving it with scattering Green's functions synthesizes a holographic scatterer indistinguishable from a target for arbitrary illuminations. Combining the two, using heterogeneous Green's functions, replaces the scattering signature of one object with that of another, transforming its acoustic identity. We demonstrate the framework in three-dimensional FDTD simulations driven by impulsive Green's functions, complemented by data-driven Green's-function retrieval, establishing a direct route to real-time 3D acoustic cloaking, holography, cloning, and disguising.

Autonomous Underwater Vehicles (AUVs) traditionally rely on complex, heavily engineered pipelines for perception, path planning, and motion control. This paper explores the feasibility of an end-to-end Deep Reinforcement Learning (DRL) approach that maps raw sensor data directly to thruster commands, reducing manual engineering. We propose a hierarchical reinforcement learning (HRL) architecture splitting the problem into two Markov Decision Processes. A High-Level (HL) policy operating at 2Hz processes raw $84 \times 84$ pixel monocular camera frames, stacked $100 \times 100$ pixel forward-looking imaging sonar, and proprioceptive data to generate spatial subgoals. Simultaneously, a Low-Level (LL) policy operating at 10Hz converts these subgoals into thruster commands. The HL policy is trained using Reinforcement Learning from Prior Demonstrations (RLPD) within a modified Sample-Efficient Robotic Reinforcement Learning (SERL) framework, while the LL policy utilizes Soft Actor-Critic (SAC) combined with Hindsight Experience Replay (HER). Evaluated in the high-fidelity HoloOcean simulator, our method demonstrates successful obstacle avoidance, achieving trajectory lengths closely approximating (within 4% to 6% of) an $\text{RRT}^*$ planning baseline. Furthermore, the learned policy exhibits strong robustness to simulated sensor noise and decreased visibility. While the system navigates familiar geometries effectively, experiments reveal generalization limitations when encountering unvisited areas with novel obstacle shapes. Ultimately, this work demonstrates the promise of sample-efficient, end-to-end DRL for underwater navigation using minimal computational hardware.