Papers

In the design of engineered components, rigorous vibration testing is essential for performance validation and identification of resonant frequencies and amplitudes encountered during operation. Performing this evaluation numerically via machine learning has great potential to accelerate design iteration and make testing workflows more efficient. However, dynamical systems are conventionally difficult to solve via machine learning methods without using physics-based regularizing loss functions. To properly perform this forecasting task, a structure that has an inspectable physical obedience can be devised without the use of regularizing terms from first principles. The method employed in this work is a neural operator integrated with an implicit numerical scheme. This architecture enables operators to learn of the underlying state-space dynamics from limited data, allowing generalization to untested driving frequencies and initial conditions. This network can infer the system's global frequency response by training on a small set of input conditions. As a foundational proof of concept, this investigation verifies the machine learning algorithm with a linear, single-degree-of-freedom system, demonstrating implicit obedience of dynamics. This approach demonstrates 99.87% accuracy in predicting the Frequency Response Curve (FRC), forecasting the frequency and amplitude of linear resonance training on 7% of the bandwidth of the solution. By training machine learning models to internalize physics information rather than trajectory, better generalization accuracy can be realized, vastly improving the timeframe for vibration studies on engineered components.

Power distribution systems are increasingly exposed to large voltage fluctuations driven by intermittent renewable generation and time varying loads (e.g., electric vehicles and storage). To address this challenge, a number of advanced controllers have been proposed for voltage regulation. However, these controllers typically rely on fixed linear approximations of voltage dynamics. As a result, the solutions may become infeasible when applied to the actual voltage behavior governed by nonlinear power flow equations, particularly under heavy power injection from distributed energy resources. This paper proposes a data-driven successive linearization approach for voltage control under nonlinear power flow constraints. By leveraging the fact that the deviation between the nonlinear power flow solution and its linearization is bounded by the distance from the operating point, we perform data-driven linearization around the most recent operating point. Convergence of the proposed method to a neighborhood of KKT points is established by exploiting the convexity of the objective function and structural properties of the nonlinear constraints. Case studies show that the proposed approach achieves fast convergence and adapts quickly to changes in net load.

This paper proves that the Epistemic Support-Point Filter (ESPF) is the unique optimal recursive estimator within the class of epistemically admissible evidence-only filters. Where Bayesian filters minimize mean squared error and are driven toward an assumed truth, the ESPF minimizes maximum entropy and surfaces what has not been proven impossible – a fundamentally different epistemic commitment with fundamentally different failure modes. Two results locate this theorem within the broader landscape of estimation theory. The first is a unification: the ESPF's optimality criterion is the log-geometric mean of the alpha-cut volume family in the Holder mean hierarchy. The Popperian minimax bound and the Kalman MMSE criterion occupy the p=+inf and p=0 positions on the same curve. Possibility and probability are not competing frameworks: they are the same ignorance functional evaluated under different alpha-cut geometries. The Kalman filter is the Gaussian specialization of the ESPF's optimality criterion, not a separate invention. The second result is a diagnostic: numerical validation over a 2-day, 877-step Smolyak Level-3 orbital tracking run shows that possibilistic stress manifests through necessity saturation and surprisal escalation rather than MVEE sign change – a direct consequence of the Holder ordering, not an empirical observation. Three lemmas establish the result: the Possibilistic Entropy Lemma decomposes the ignorance functional; the Possibilistic Cramer-Rao Bound limits entropy reduction per measurement; the Evidence-Optimality Lemma proves minimum-q selection is the unique minimizer and that any rule incorporating prior possibility risks race-to-bottom bias.

A central question in modern deep learning is how to design optimizers whose behavior remains stable as the network width $w$ increases. We address this question by interpreting several widely used neural-network optimizers, including AdamW and Muon, as instances of steepest descent under matrix operator norms. This perspective links optimizer geometry with the Lipschitz structure of the network forward map, and enables width-independent control of both Lipschitz and smoothness constants. However, steepest-descent rules induced by standard $p \to q$ operator norms lack layerwise composability and therefore cannot provide width-independent bounds in deep architectures. We overcome this limitation by introducing a family of mean-normalized operator norms, denoted $\pmean \to \qmean$, that admit layerwise composability, yield width-independent smoothness bounds, and give rise to practical optimizers such as rescaled AdamW, row normalization, and column normalization. The resulting learning rate width-aware scaling rules recover $\mu$P scaling~\cite{yang2021tensor} as a special case and provide a principled mechanism for cross-width learning-rate transfer across a broad class of optimizers. We further show that Muon can suffer an $\mathcal{O}(\sqrt{w})$ worst-case growth in the smoothness constant, whereas a new family of row-normalized optimizers we propose achieves width-independent smoothness guarantees. Based on the observations, we propose MOGA (Matrix Operator Geometry Aware), a width-aware optimizer based only on row/column-wise normalization that enables stable learning-rate transfer across model widths. Large-scale pre-training on GPT-2 and LLaMA shows that MOGA, especially with row normalization, is competitive with Muon while being notably faster in large-token and low-loss regimes.

In the diverse landscape of 6G networks, where wireless connectivity demands surge and spectrum resources remain limited, flexible spectrum access becomes paramount. The success of crafting such schemes hinges on our ability to accurately characterize spectrum demand patterns across space and time. This paper presents a data-driven methodology for estimating spectrum demand variations over space and identifying key drivers of these variations in the mobile broadband landscape. By leveraging geospatial analytics and machine learning, the methodology is applied to a case study in Canada to estimate spectrum demand dynamics in urban regions. Our proposed model captures 70% of the variability in spectrum demand when trained on one urban area and tested on another. These insights empower regulators to navigate the complexities of 6G networks and devise effective policies to meet future network demands.

The progressive electrification of aircraft systems under the more electric aircraft (MEA) paradigm is reshaping the design and qualification constraints of safety-critical avionics. Emergency locator transmitters (ELTs), which are essential for post-accident localization and search and rescue (SAR) operations, have evolved from legacy 121.5/243 MHz beacons to digitally encoded 406 MHz systems, typically retaining 121.5 MHz as a homing signal in combined units. In parallel, the modernization of the Cospas-Sarsat infrastructure, especially MEOSAR, together with multi-constellation global navigation satellite system (GNSS) integration and second-generation beacon capabilities, is reducing detection latency and enabling richer distress messaging. However, MEA platforms impose stricter constraints on available power, thermal management, wiring density, and electromagnetic compatibility (EMC). As a result, ELT performance increasingly depends not only on the device itself, but also on its installation conditions and on the aircraft's overall electrical environment. This review summarizes the ELT architectures and activation/operational cycles, outlines key technological milestones, and consolidates the main integration challenges for MEA, with emphasis on energy autonomy, battery qualification frameworks, EMC and installation practices, and survivability-driven failure modes (e.g., antenna/feedline damage, mounting, and post-impact shielding). Finally, emerging trends include ELT for distress tracking (DT), energy-based designs, advanced health monitoring, and certification-ready pathways for next-generation SAR services are discussed, highlighting research directions that can deliver demonstrable, certifiable gains in reliability, energy efficiency, and robust integration for future electrified aircraft.

Accurately forecasting spectrum demand is a key component for efficient spectrum resource allocation and management. With the rapid growth in demand for wireless services, mobile network operators and regulators face increasing challenges in ensuring adequate spectrum availability. This paper presents a data-driven approach leveraging artificial intelligence (AI) and machine learning (ML) to estimate and manage spectrum demand. The approach uses multiple proxies of spectrum demand, drawing from site license data and derived from crowdsourced data. These proxies are validated against real-world mobile network traffic data to ensure reliability, achieving an R$^2$ value of 0.89 for an enhanced proxy. The proposed ML models are tested and validated across five major Canadian cities, demonstrating their generalizability and robustness. These contributions assist spectrum regulators in dynamic spectrum planning, enabling better resource allocation and policy adjustments to meet future network demands.

Existing aerial-robotics benchmarks target vehicles from hundreds of grams to several kilograms and typically expose only high-level state data. They omit the actuator-level signals required to study nano-scale quadrotors, where low-Reynoldsnumber aerodynamics, coreless DC motor nonlinearities, and severe computational constraints invalidate models and controllers developed for larger vehicles. We introduce NanoBench, an open-source multi-task benchmark collected on the commercially available Crazyflie 2.1 nano-quadrotor (takeoff weight 27 g) in a Vicon motion capture arena. The dataset comprises 172 flight recordings spanning multi-frequency excitation, geometric trajectory tracking at three speed regimes, and long-duration battery-drain hover. Each recording provides synchronized Vicon ground truth, raw IMU data, onboard extended Kalman filter (EKF) estimates, PID controller internals, and motor PWM commands at 100 Hz, alongside battery telemetry at 10 Hz, crossaligned by gyroscope-based correlation on a 1 ms search grid. NanoBench defines standardized evaluation protocols, train/test splits, and open-source baselines for three tasks: nonlinear system identification, closed-loop controller benchmarking, and onboard state estimation assessment. To our knowledge, it is the first public dataset to jointly provide actuator commands, controller internals, and estimator outputs with millimeter-accurate ground truth on a commercially available nano-scale aerial platform.

A privacy-preserving dynamic average consensus (DAC) algorithm is proposed that achieves consensus while preventing external eavesdroppers from inferring the reference signals and their derivatives. During the initialization phase, each agent generates a set of sinusoidal signals with randomly selected frequencies and exchanges them with its neighboring agents to construct a masking signal. Each agent masks its reference signals using this composite masking signal before executing the DAC update rule. It is shown that the developed scheme preserves the convergence properties of the conventional DAC framework while preventing information leakage to external eavesdroppers. Furthermore, the developed algorithm is applied to state-of-charge (SoC) balancing in a networked battery energy storage system to demonstrate its practical applicability. Simulation results validate the theoretical findings.

The next generation of cellular networks is designed to provide ubiquitous connectivity to a wide range of devices. As Telecommunication Service Providers (TSPs) increasingly collaborate with public cloud providers to deploy 5G and beyond networks, a fundamental shift is underway, from hardware-bound Physical Network Functions (PNFs) to cloud-native, containerized deployments managed through platforms like Kubernetes. While this transition promises greater scalability, flexibility, and cost efficiency, it also introduces a complex set of technical and operational challenges that must be thoroughly understood before large-scale cellular deployments can take place in cloud environments. In this survey, we present a structured taxonomy that categorizes the design space of cloud-based cellular deployments across four dimensions: deployment architecture, resource management and orchestration, multi-tenancy and isolation, and economic and ownership models. Using this taxonomy as a foundation, we critically analyze six key investigation areas, security and privacy, scalability and elasticity, performance and latency, cost optimization, resilience and fault management, and compliance and sovereignty, examining each through a cloud-native lens. To benchmark the state of industry adoption, we examine the deployment strategies of leading Infrastructure-as-a-Service (IaaS) providers, namely Amazon Web Services (AWS), Microsoft Azure, and Google Cloud Platform (GCP). Finally, we identify emerging trends such as AI-driven orchestration, quantum-safe protocols for virtualized network functions, and serverless networking for 6G, while articulating the open challenges that remain in realizing robust, scalable cloud-based cellular networks.

This paper proposes a posterior sampling-based beam training framework for near-field communication under multi-path channels. By leveraging Thompson Sampling (TS), the framework adaptively balances exploration and exploitation to maximize the final beamforming gain under a finite pilot overhead. To ensure data-efficient learning, we incorporate a structured Gaussian prior in the DFT domain and use an RBF covariance as a tractable local-correlation prior, motivated by near-field energy leakage, to promote information sharing among neighboring DFT components. We develop three TS strategies: codebook-constrained search for rapid stabilization via structural regularization, continuous-space search for high accuracy refinement, and a two-stage hybrid refinement scheme that balances training efficiency and estimation accuracy. Simulation results show that the proposed framework reduces pilot overhead by over 90% while achieving more than a 2 dB SNR gain over baselines in multipath environments. Furthermore, simulations show that the continuous-space search approaches the full-CSI benchmark as the available pilot numbers become sufficiently large.

A 3 bit Analog to Digital Converter (ADC) is designed using perpendicular Spin Orbit Torque Magnetic Tunnel Junction (SOT MTJ). A sampled analog input signal is transmitted as a spin orbit torque current (Iin) to a perpendicular SOT MTJ, and deterministic switching is supported by the Voltage Controlled Magnetic Anisotropy (VCMA) and Spin Transfer Torque (STT) switching methods. Analog to digital conversion is done by comparing input signal with varied critical current of SOT MTJs. The critical current of each is SOT MTJ governed by varying widths of Heavy Metal (HM). In the 3 bit ADC, there are two sets of 7 SOT MTJs for quantizing input value, a conversion set and dummy set for comparing the change in resistance state. As input signal passed through conversion set SOT MTJs switches from Parallel (P) to AntiParallel (AP) state if the input signal exceeds its critical current. The conversion set change in state is converted to thermometer codes by StrongARM latch comparator by comparing the resistance with dummy set SOT MTJs, where all the in P state or low resistance. A novel architecture is proposed for increasing speed of throughput, by utilizing the dummy set of as a conversion set and conversion set as dummy set, thus eliminating the reset step from analog to digital conversion. And by improving SOT-MTJ and timing blocks a field free spin flash ADC has a power consumption of 476 uW with a conversion rate of 304.1 MHz is produced.

The surge in wireless connectivity demand, coupled with the finite nature of spectrum resources, compels the development of efficient spectrum management approaches. Spectrum sharing presents a promising avenue, although it demands precise characterization of spectrum demand for informed policy-making. This paper introduces HR-GAT, a hierarchical resolution graph attention network model, designed to predict spectrum demand using geospatial data. HR-GAT adeptly handles complex spatial demand patterns and resolves issues of spatial autocorrelation that usually challenge standard machine learning models, often resulting in poor generalization. Tested across five major Canadian cities, HR-GAT improves predictive accuracy of spectrum demand by 21% over eight baseline models, underscoring its superior performance and reliability.

Super-resolution ultrasound via microbubble (MB) localisation and tracking, also known as ultrasound localisation microscopy (ULM), can resolve microvasculature beyond the acoustic diffraction limit. However, significant challenges remain in localisation performance and data acquisition and processing time. Deep learning methods for ULM have shown promise to address these challenges, however, they remain limited by in vivo label scarcity and the simulation-to-reality domain gap. We present CycleULM, the first unified label-free deep learning framework for ULM. CycleULM learns a physics-emulating translation between the real contrast-enhanced ultrasound (CEUS) data domain and a simplified MB-only domain, leveraging the power of CycleGAN without requiring paired ground truth data. With this translation, CycleULM removes dependence on high-fidelity simulators or labelled data, and makes MB localisation and tracking substantially easier. Deployed as modular plug-and-play components within existing pipelines or as an end-to-end processing framework, CycleULM delivers substantial performance gains across both in silico and in vivo datasets. Specifically, CycleULM improves image contrast (contrast-to-noise ratio) by up to 15.3 dB and sharpens CEUS resolution with a 2.5{\times} reduction in the full width at half maximum of the point spread function. CycleULM also improves MB localisation performance, with up to +40% recall, +46% precision, and a -14.0 {\mu}m mean localisation error, yielding more faithful vascular reconstructions. Importantly, CycleULM achieves real-time processing throughput at 18.3 frames per second with order-of-magnitude speed-ups (up to ~14.5{\times}). By combining label-free learning, performance enhancement, and computational efficiency, CycleULM provides a practical pathway toward robust, real-time ULM and accelerates its translation to clinical applications.

Frequency stability is fundamental to the secure operation of power systems. With growing uncertainty and volatility introduced by renewable generation, secondary frequency regulation must now deliver enhanced performance not only in the steady state but also during transients. This paper presents a systematic framework to embed learning in the design of a primal-dual controller that provides provable (potentially exponential) stability and steady-state optimality, while simultaneously improving key transient metrics, including frequency nadir and control effort, in a data-driven manner. In particular, we employ the primal-dual dynamics of an optimization problem that encodes steady-state objectives to realize secondary frequency control with asymptotic stability guarantee. To augment transient performance of the controller via learning, a change of variables on control inputs, which will be deployed by neural networks, is proposed such that under mild conditions, stability and steady-state optimality are preserved. It further allows us to define a learning goal that accounts for the exponential convergence rate, frequency nadir and accumulated control effort, and use sample trajectories to enhance these metrics. Simulation results validate the theories and demonstrate superior transient performance of the learning-augmented primal-dual controller.

Accurate path loss prediction is crucial for wireless network planning and optimization in suburban environments with complex terrain variation and diverse land cover. This paper proposes a model assisted hybrid path loss prediction method that introduces an environment adaptive compensation on top of the classic close-in free-space reference distance (CI) path loss model. By jointly predicting the path loss exponent and a compensation term, the proposed approach dynamically adjusts the empirical trend. To improve the effectiveness of environmental representation, three environmental image organization schemes are constructed and evaluated. Experiments on measurement data collected in Pingtan Island show that the proposed method outperforms the CI model and a conventional model assisted baseline, achieving a test root mean square error of 4.04 dB.

Terahertz (THz) radiation provides a non-ionizing, highly sensitive probe of the dielectric properties of biological tissues. In this study, we present a comprehensive experimental characterization of dielectric properties using pork skin tissue, a widely used surrogate for human tissue, as a biological sample. Measurements are conducted employing THz time-domain spectroscopy in the 0.1-11 THz frequency range with photoconductive antennas for both signal generation and detection. Frequency-dependent refractive indices, absorption, and complex permittivity are extracted from transmitted time-domain signals. Our results confirm strong absorption and low transmittance at low THz frequencies due to water content, while highlighting frequency-dependent dispersion and narrowband transmission features at higher frequencies. This work provides one of the first extended-frequency datasets of biological tissue dielectric properties, supporting realistic channel modeling for the design and development of intra-body nanosensor networks in the THz band.

This study presents a comprehensive experimental assessment of a low-cost frequency-modulated continuous-wave (FMCW) multiple-input multiple-output (MIMO) radar for non-contact vital sign monitoring, focusing on respiratory rate (RR) and heart rate (HR) estimation. The influence of sensing distance and number of transmitted chirps on measurement accuracy is systematically quantified. Results exhibit a U-shaped error profile with optimal performance near $70~cm$, achieving mean absolute errors of $0.8~bpm$ for RR and $3.2~bpm$ for HR. Accuracy deteriorates at short ($<60~cm$) and long ($>100~cm$) distances due to multipath, near-field, and signal-to-noise effects. Increasing chirp count enhances performance: RR errors converge asymptotically for $\geq96$ chirps, while HR requires at least 96 chirps for stable detection. Variability metrics, including heart and respiratory rate variability, remain less accurate ($>15$–$30\%$ error), indicating limited capability in capturing instantaneous fluctuations. These findings define a fundamental trade-off: the radar ensures robust estimation of average RR and HR but exhibits restricted precision in high-resolution beat-to-beat and breath-to-breath monitoring.

Nonlinear optimisation techniques are commonly employed to minimise complex cost functions, with their effectiveness determined largely by the structure of the underlying error landscape. These methods require initial parameter values, and in the presence of multiple local minima, they are prone to becoming trapped in suboptimal regions. The likelihood of locating the global minimum increases substantially when the initialisation lies within its corresponding basin of attraction. Consequently, high-quality initial parameters are critical for successful optimisation. This technical report outlines a new strategy for selecting suitable initial parameters for a trigonometric model and unevenly sampled data, ensuring that the optimisation procedure starts sufficiently close to the global minimum. The proposed parameter estimation approach is strictly NI-based, interpretable, and explainable. It targets at complicated cases which include: samples with strong random noise, samples with only few covered periods, and samples which cover only a fraction of one period. Special attention is put on the frequency estimation. It can be shown that an estimation of initial parameters with sufficient accuracy is possible down to a signal-noise-ratio of 1.4 dB at much lower computational costs than the Lomb-Scargle-periodogram method requires.

Accurate relative positioning is crucial for swarm aerial robotics, enabling coordinated flight and collision avoidance. Although vision-based tracking has been extensively studied, 3D LiDAR-based methods remain underutilized despite their robustness under varying lighting conditions. Existing systems often rely on bulky, power-intensive sensors, making them impractical for small UAVs with strict payload and energy constraints. This paper presents a lightweight LiDAR-based UAV tracking system incorporating an Adaptive Extended Kalman Filter (AEKF) framework. Our approach effectively addresses the challenges posed by sparse, noisy, and nonuniform point cloud data generated by non-repetitive scanning 3D LiDARs, ensuring reliable tracking while remaining suitable for small drones with strict payload constraints. Unlike conventional filtering techniques, the proposed method dynamically adjusts the noise covariance matrices using innovation and residual statistics, thereby enhancing tracking accuracy under real-world conditions. Additionally, a recovery mechanism ensures continuity of tracking during temporary detection failures caused by scattered LiDAR returns or occlusions. Experimental validation was performed using a Livox Mid-360 LiDAR mounted on a DJI F550 UAV in real-world flight scenarios. The proposed method demonstrated robust UAV tracking performance under sparse LiDAR returns and intermittent detections, consistently outperforming both standard Kalman filtering and particle filtering approaches during aggressive maneuvers. These results confirm that the framework enables reliable relative positioning in GPS-denied environments without the need for multi-sensor arrays or external infrastructure.

Global perception is essential for embodied agents in 360{\deg} spaces, yet current affordance grounding remains largely object-centric and restricted to perspective views. To bridge this gap, we introduce a novel task: Holistic Affordance Grounding in 360{\deg} Indoor Environments. This task faces unique challenges, including severe geometric distortions from Equirectangular Projection (ERP), semantic dispersion, and cross-scale alignment difficulties. We propose PanoAffordanceNet, an end-to-end framework featuring a Distortion-Aware Spectral Modulator (DASM) for latitude-dependent calibration and an Omni-Spherical Densification Head (OSDH) to restore topological continuity from sparse activations. By integrating multi-level constraints comprising pixel-wise, distributional, and region-text contrastive objectives, our framework effectively suppresses semantic drift under low supervision. Furthermore, we construct 360-AGD, the first high-quality panoramic affordance grounding dataset. Extensive experiments demonstrate that PanoAffordanceNet significantly outperforms existing methods, establishing a solid baseline for scene-level perception in embodied intelligence. The source code and benchmark dataset will be made publicly available at https://github.com/GL-ZHU925/PanoAffordanceNet.

Semantic occupancy prediction enables dense 3D geometric and semantic understanding for autonomous driving. However, existing camera-based approaches implicitly assume complete surround-view observations, an assumption that rarely holds in real-world deployment due to occlusion, hardware malfunction, or communication failures. We study semantic occupancy prediction under incomplete multi-camera inputs and introduce $M^2$-Occ, a framework designed to preserve geometric structure and semantic coherence when views are missing. $M^2$-Occ addresses two complementary challenges. First, a Multi-view Masked Reconstruction (MMR) module leverages the spatial overlap among neighboring cameras to recover missing-view representations directly in the feature space. Second, a Feature Memory Module (FMM) introduces a learnable memory bank that stores class-level semantic prototypes. By retrieving and integrating these global priors, the FMM refines ambiguous voxel features, ensuring semantic consistency even when observational evidence is incomplete. We introduce a systematic missing-view evaluation protocol on the nuScenes-based SurroundOcc benchmark, encompassing both deterministic single-view failures and stochastic multi-view dropout scenarios. Under the safety-critical missing back-view setting, $M^2$-Occ improves the IoU by 4.36%. As the number of missing cameras increases, the robustness gap further widens; for instance, under the setting with five missing views, our method boosts the IoU by 6.67%. These gains are achieved without compromising full-view performance. The source code will be publicly released at https://github.com/qixi7up/M2-Occ.

[This work has been submitted to the IEEE for possible publication. Copyright may be transferred without notice, after which this version may no longer be accessible.] In a wireless acoustic sensor network (WASN), devices (i.e., nodes) can collaborate through distributed algorithms to collectively perform audio signal processing tasks. This paper focuses on the distributed estimation of node-specific desired speech signals using network-wide Wiener filtering. The objective is to match the performance of a centralized system that would have access to all microphone signals, while reducing the communication bandwidth usage of the algorithm. Existing solutions, such as the distributed adaptive node-specific signal estimation (DANSE) algorithm, converge towards the multichannel Wiener filter (MWF) which solves a centralized linear minimum mean square error (LMMSE) signal estimation problem. However, they do so iteratively, which can be slow and impractical. Many solutions also assume that all nodes observe the same set of sources of interest, which is often not the case in practice. To overcome these limitations, we propose the distributed multichannel Wiener filter (dMWF) for fully connected WASNs. The dMWF is non-iterative and optimal even when nodes observe different sets of sources. In this algorithm, nodes exchange neighbor-pair-specific, low-dimensional (fused) signals estimating the contribution of sources observed by both nodes in the pair. We formally prove the optimality of dMWF and demonstrate its performance in simulated speech enhancement experiments. The proposed algorithm is shown to outperform DANSE in terms of objective metrics after short operation times, highlighting the benefit of its iterationless design.

Animal brains exhibit remarkable efficiency in perception and action, while being robust to both external and internal perturbations. The means by which brains accomplish this remains, for now, poorly understood, hindering our understanding of animal and human cognition, as well as our own implementation of efficient algorithms for control of dynamical systems.A potential candidate for a robust mechanism of state estimation and action computation is the free energy principle, but existing implementations of this principle have largely relied on conventional, biologically implausible approaches without spikes. We propose a novel, efficient, and robust spiking control framework with realistic biological characteristics. The resulting networks function as free energy constrainers, in which neurons only fire if they reduce the free energy of their internal representation. The networks offer efficient operation through highly sparse activity while matching performance with other similar spiking frameworks, and have high resilience against both external (e.g. sensory noise or collisions) and internal perturbations (e.g. synaptic noise and delays or neuron silencing) that such a network would be faced with when deployed by either an organism or an engineer. Overall, our work provides a novel mathematical account for spiking control through constraining free energy, providing both better insight into how brain networks might leverage their spiking substrate and a new route for implementing efficient control algorithms in neuromorphic hardware.

We present DRES: a 1.5-hour Dutch realistic elicited (semi-spontaneous) speech dataset from 80 speakers recorded in noisy, public indoor environments. DRES was designed as a test set for the evaluation of state-of-the-art (SotA) automatic speech recognition (ASR) and speech enhancement (SE) models in a real-world scenario: a person speaking in a public indoor space with background talkers and noise. The speech was recorded with a four-channel linear microphone array. In this work we evaluate the speech quality of five well-known single-channel SE algorithms and the recognition performance of eight SotA off-the-shelf ASR models before and after applying SE on the speech of DRES. We found that five out of the eight ASR models have WERs lower than 22% on DRES, despite the challenging conditions. In contrast to recent work, we did not find a positive effect of modern single-channel SE on ASR performance, emphasizing the importance of evaluating in realistic conditions.

While multi-audio understanding is critical for large audio-language models (LALMs), it remains underexplored. We introduce MUGEN, a comprehensive benchmark evaluating this capability across speech, general audio, and music. Our experiments reveal consistent weaknesses in multi-audio settings, and performance degrades sharply as the number of concurrent audio inputs increases, identifying input scaling as a fundamental bottleneck. We further investigate training-free strategies and observe that Audio-Permutational Self-Consistency, which diversifies the order of audio candidates, helps models form more robust aggregated predictions, yielding up to 6.28% accuracy gains. Combining this permutation strategy with Chain-of-Thought further improves performance to 6.74%. These results expose blind spots in current LALMs and provide a foundation for evaluating complex auditory comprehension.

Room Impulse Responses (RIRs) enable realistic acoustic simulation, with applications ranging from multimedia production to speech data augmentation. However, acquiring high-quality real-world RIRs is labor-intensive, and data scarcity remains a challenge for data-driven RIR generation approaches. In this paper, we propose a novel approach to RIR generation by adapting a pre-trained text-to-audio model, demonstrating for the first time that large-scale generative audio priors can be effectively leveraged for this task. To address the lack of text-RIR paired data, we utilize a labeling pipeline leveraging vision-language models to extract acoustic descriptions from existing image-RIR datasets. We introduce an in-context learning strategy to accommodate free-form user prompts during inference. Evaluations including a subjective listening test demonstrate that our model generates plausible RIRs with substantially less training data. Audio examples are available on our demo website.

Current developments of high-speed magnetic levitation technology using the principle of the electromagnet suspension (EMS) focus on reaching vehicle speeds of more than 600 km/h. With increasing vehicle speeds, however, updated control algorithms need to be investigated to reliably stabilize the system and meet the demands in terms of ride comfort. This article examines the modern and popular approach of model predictive control and its application to the magnetic levitation control system. Investigated key aspects are the parameterization of the model predictive controller and its implementation on embedded, resource constrained hardware. The results reveal that model predictive control is capable to robustly stabilize the highly nonlinear and constrained system even at very high speed. Furthermore, processor-in-the-loop studies are carried out to validate the designed control algorithms on a microcontroller.

Maintaining background consistency while enhancing foreground quality remains a core challenge in video editing. Injecting full-image information often leads to background artifacts, whereas rigid background locking severely constrains the model's capacity for foreground generation. To address this issue, we propose KV-Lock, a training-free framework tailored for DiT-based video diffusion models. Our core insight is that the hallucination metric (variance of denoising prediction) directly quantifies generation diversity, which is inherently linked to the classifier-free guidance (CFG) scale. Building upon this, KV-Lock leverages diffusion hallucination detection to dynamically schedule two key components: the fusion ratio between cached background key-values (KVs) and newly generated KVs, and the CFG scale. When hallucination risk is detected, KV-Lock strengthens background KV locking and simultaneously amplifies conditional guidance for foreground generation, thereby mitigating artifacts and improving generation fidelity. As a training-free, plug-and-play module, KV-Lock can be easily integrated into any pre-trained DiT-based models. Extensive experiments validate that our method outperforms existing approaches in improved foreground quality with high background fidelity across various video editing tasks.

Finetuning wireless receivers to a specific deployment scenario can yield significant error-rate performance improvements without increasing processing complexity. However, site-specific finetuning has so far only been demonstrated on synthetic channel data and lacks real-world benchmarks. In this work, we empirically study site-specific finetuning of neural receivers using real-world 5G NR physical uplink shared channel (PUSCH) data collected with an over-the-air testbed at ETH Zurich across three scenarios: (i) a small laboratory, (ii) a large office floor, and (iii) a high-mobility outdoor environment. Our results confirm substantial error-rate performance improvements from site-specific finetuning, consistent with earlier findings based on synthetic channel data. Moreover, we demonstrate that these improvements generalize across different user-equipment hardware and deployment scenarios.