Papers

We propose a method to efficiently compute tomographic projections of a 3D volume represented by a linear combination of shifted B-splines. To do so, we propose a ray-tracing algorithm that computes 3D line integrals with arbitrary projection geometries. One of the components of our algorithm is a neural network that computes the contribution of the basis functions efficiently. In our experiments, we consider well-posed cases where the data are sufficient for accurate reconstruction without the need for regularization. We achieve higher reconstruction quality than traditional voxel-based methods.

Neural Video Representation~(NVR) is a promising paradigm for video compression, showing great potential in improving video storage and transmission efficiency. While recent advances have made efforts in architectural refinements to improve representational capability, these methods typically involve complex designs, which may incur increased computational overhead and lack the flexibility to integrate into other frameworks. Moreover, the inherent limitation in model capacity restricts the expressiveness of NVR networks, resulting in a performance bottleneck. To overcome these limitations, we propose Online-RepNeRV, a NVR framework based on online structural reparameterization. Specifically, we propose a universal reparameterization block named ERB, which incorporates multiple parallel convolutional paths to enhance the model capacity. To mitigate the overhead, an online reparameterization strategy is adopted to dynamically fuse the parameters during training, and the multi-branch structure is equivalently converted into a single-branch structure after training. As a result, the additional computational and parameter complexity is confined to the encoding stage, without affecting the decoding efficiency. Extensive experiments on mainstream video datasets demonstrate that our method achieves an average PSNR gain of 0.37-2.7 dB over baseline methods, while maintaining comparable training time and decoding speed.

Understanding how sustainable behaviors spread within heterogeneous societies requires the integration of behavioral data, social influence mechanisms, and structured approaches to control. In this paper, we propose a data-driven computational framework for coupled opinion-adoption dynamics in social systems. Each node in the multilayer network represents a community characterized by a specific age group and mobility level, derived from large-scale survey data on the predisposition to adopt electric vehicles in Northern Europe. The proposed model captures three mechanisms: behavioral contagion through social and informational diffusion, abandonment driven by dissatisfaction, and feedback between opinions and adoption levels through social influence. Analyzing the equilibrium points of the coupled system allows us to derive the conditions that enable large-scale adoption. We empirically calibrate the model using data to construct synthetic populations and social similarity networks, which we use to explore targeted interventions that promote sustainable transitions. Specifically, the analysis focuses on two types of control strategies: opinion-based policies, which act on the social network layer, and policies that aim to improve experience and reduce dissatisfaction. Simulation results show that the latter ensure more stable and long-term adoption, offering concrete insights for designing effective interventions in sociotechnical transitions toward sustainability.

Estimating the Region of Attraction (RoA) for nonlinear dynamical systems is a fundamental problem in control theory, with direct implications for stability analysis and safe controller design. Traditional approaches rely on analytically derived Lyapunov functions, which are often conservative and challenging to construct for high-dimensional or highly nonlinear systems. In this work, we propose a data-driven framework for learning and verifying RoA estimates for nonlinear systems using neural-network-based Lyapunov functions. Our method employs a composite Lyapunov function that combines a quadratic term with a neural-network-based component, providing both structure and flexibility. We introduce a novel homogeneous loss function for training, which removes the imbalance typically caused by the two non-homogeneous Lyapunov conditions. Together, these two aspects enable efficient training of the Lyapunov candidate. To guarantee the correctness of the learned Lyapunov function, we employ a Satisfiability Modulo Theories (SMT) solver to formally verify the stability results. Lastly, we perform a deeper analysis near the origin to overcome numerical artifacts, ensuring strict asymptotic stability. We demonstrate the effectiveness of our approach on benchmark nonlinear systems, showing that it significantly reduces conservatism compared to traditional Lyapunov methods while maintaining verifiability. This framework bridges the gap between function approximation and stability certification, paving the way for scalable safety analysis in learning-based control and safety-critical applications.

The rapid development of sixth-generation (6G) wireless networks requires seamless integration of communication and sensing to support ubiquitous intelligence and real-time, high-reliability applications. Integrated sensing and communication (ISAC) has emerged as a key solution for achieving this convergence, offering joint utilization of spectral, hardware, and computing resources. However, realizing high-performance ISAC remains challenging due to environmental line-of-sight (LoS) blockage, limited spatial resolution, and the inherent coverage asymmetry and resource coupling between sensing and communication. Intelligent reflecting surfaces (IRSs), featuring low-cost, energy-efficient, and programmable electromagnetic reconfiguration, provide a promising solution to overcome these limitations. This article presents a comprehensive overview of IRS-aided wireless sensing and ISAC technologies, including IRS architectures, target detection and estimation techniques, beamforming designs, and performance metrics. It further explores IRS-enabled new opportunities for more efficient performance balancing, coexistence, and networking in ISAC systems, focuses on current design bottlenecks, and outlines future research directions. This article aims to offer a unified design framework that guides the development of practical and scalable IRS-aided ISAC systems for the next-generation wireless network.

Correlating articular cartilage material properties to quantitative magnetic resonance imaging biomarkers is a powerful approach to biofidelic finite element models. However, subject-specific relationships between imaging biomarkers such as T2 and material properties like dynamic modulus are uncertain. To evaluate the sensitivity of finite element models to this uncertainty, we shifted the slope and intercept of a linear T2-dynamic modulus relationship used to define cartilage properties. Modulus shifts led to notable percent changes in the top 1% of calculated stress and strain while modulating slope had a negligible impact, together supporting the use of physiologically relevant moduli ranges in subject-specific models.

Modern text-to-speech (TTS) systems, particularly those built on Large Audio-Language Models (LALMs), generate high-fidelity speech that faithfully reproduces input text and mimics specified speaker identities. While prior misuse studies have focused on speaker impersonation, this work explores a distinct content-centric threat: exploiting TTS systems to produce speech containing harmful content. Realizing such threats poses two core challenges: (1) LALM safety alignment frequently rejects harmful prompts, yet existing jailbreak attacks are ill-suited for TTS because these systems are designed to faithfully vocalize any input text, and (2) real-world deployment pipelines often employ input/output filters that block harmful text and audio. We present HARMGEN, a suite of five attacks organized into two families that address these challenges. The first family employs semantic obfuscation techniques (Concat, Shuffle) that conceal harmful content within text. The second leverages audio-modality exploits (Read, Spell, Phoneme) that inject harmful content through auxiliary audio channels while maintaining benign textual prompts. Through evaluation across five commercial LALMs-based TTS systems and three datasets spanning two languages, we demonstrate that our attacks substantially reduce refusal rates and increase the toxicity of generated speech. We further assess both reactive countermeasures deployed by audio-streaming platforms and proactive defenses implemented by TTS providers. Our analysis reveals critical vulnerabilities: deepfake detectors underperform on high-fidelity audio; reactive moderation can be circumvented by adversarial perturbations; while proactive moderation detects 57-93% of attacks. Our work highlights a previously underexplored content-centric misuse vector for TTS and underscore the need for robust cross-modal safeguards throughout training and deployment.

Integrated sensing and communications (ISAC) is a disruptive technology enabling future sixth-generation (6G) networks. This paper investigates target detection in a bistatic ISAC system, in which the base station (BS) transmits superimposed ISAC signals comprising both Gaussian information-bearing and deterministic sensing components to simultaneously provide communication and sensing functionalities. First, we develop a Neyman-Pearson (NP)-based detector that effectively utilizes both the deterministic sensing and random communication signals. Closed-form analysis reveals that both signal components contribute to improving the overall detection performance. Subsequently, we optimize the BS transmit beamforming to maximize the detection probability, subject to a minimum signal-to-interference-plus-noise ratio (SINR) constraint for the communication user (CU) and a total transmit power budget at the BS. The resulting non-convex beamforming optimization problem is addressed via semi-definite relaxation (SDR) and successive convex approximation (SCA) techniques. Simulation results demonstrate the superiority of the proposed NP-based detector, which leverages both types of signals, over benchmark schemes that treat information signals as interference. They also reveal that a higher communication-rate threshold directs more transmit power to Gaussian information-bearing signals, thereby diminishing deterministic-signal power and weakening detection performance.

Despite remarkable advancements in supervised pansharpening neural networks, these methods face domain adaptation challenges of resolution due to the intrinsic disparity between simulated reduced-resolution training data and real-world full-resolution scenarios.To bridge this gap, we propose an unsupervised pansharpening framework, CLIPPan, that enables model training at full resolution directly by taking CLIP, a visual-language model, as a supervisor. However, directly applying CLIP to supervise pansharpening remains challenging due to its inherent bias toward natural images and limited understanding of pansharpening tasks. Therefore, we first introduce a lightweight fine-tuning pipeline that adapts CLIP to recognize low-resolution multispectral, panchromatic, and high-resolution multispectral images, as well as to understand the pansharpening process. Then, building on the adapted CLIP, we formulate a novel loss integrating semantic language constraints, which aligns image-level fusion transitions with protocol-aligned textual prompts (e.g., Wald's or Khan's descriptions), thus enabling CLIPPan to use language as a powerful supervisory signal and guide fusion learning without ground truth. Extensive experiments demonstrate that CLIPPan consistently improves spectral and spatial fidelity across various pansharpening backbones on real-world datasets, setting a new state of the art for unsupervised full-resolution pansharpening.

Time-efficient estimation of muscle activations and forces across multi-joint systems is critical for clinical assessment and assistive device control. However, conventional approaches are computationally expensive and lack a high-quality labeled dataset for multi-joint applications. To address these challenges, we propose a physics-informed deep learning framework that estimates muscle activations and forces directly from kinematics. The framework employs a novel Multi-Joint Cross-Attention (MJCA) module with Bidirectional Gated Recurrent Unit (BiGRU) layers to capture inter-joint coordination, enabling each joint to adaptively integrate motion information from others. By embedding multi-joint dynamics, inter-joint coupling, and external force interactions into the loss function, our Physics-Informed MJCA-BiGRU (PI-MJCA-BiGRU) delivers physiologically consistent predictions without labeled data while enabling time-efficient inference. Experimental validation on two datasets demonstrates that PI-MJCA-BiGRU achieves performance comparable to conventional supervised methods without requiring ground-truth labels, while the MJCA module significantly enhances inter-joint coordination modeling compared to other baseline architectures.

Quantifying greenhouse-gases (GHG) in wetlands is critical for climate modeling and restoration assessment, yet manual sampling is labor-intensive, and time demanding. We present WetExplorer, an autonomous tracked robot that automates the full GHG-sampling workflow. The robot system integrates low-ground-pressure locomotion, centimeter-accurate lift placement, dual-RTK sensor fusion, obstacle avoidance planning, and deep-learning perception in a containerized ROS2 stack. Outdoor trials verified that the sensor-fusion stack maintains a mean localization error of 1.71 cm, the vision module estimates object pose with 7 mm translational and 3{\deg} rotational accuracy, while indoor trials demonstrated that the full motion-planning pipeline positions the sampling chamber within a global tolerance of 70 mm while avoiding obstacles, all without human intervention. By eliminating the manual bottleneck, WetExplorer enables high-frequency, multi-site GHG measurements and opens the door for dense, long-duration datasets in saturated wetland terrain.

Satellite-based radar retrieval methods are widely employed to fill coverage gaps in ground-based radar systems, especially in remote areas affected by terrain blockage and limited detection range. Existing methods predominantly rely on overly simplistic spatial-domain architectures constructed from a single data source, limiting their ability to accurately capture complex precipitation patterns and sharply defined meteorological boundaries. To address these limitations, we propose WaveC2R, a novel wavelet-driven coarse-to-refined framework for radar retrieval. WaveC2R integrates complementary multi-source data and leverages frequency-domain decomposition to separately model low-frequency components for capturing precipitation patterns and high-frequency components for delineating sharply defined meteorological boundaries. Specifically, WaveC2R consists of two stages (i)Intensity-Boundary Decoupled Learning, which leverages wavelet decomposition and frequency-specific loss functions to separately optimize low-frequency intensity and high-frequency boundaries; and (ii)Detail-Enhanced Diffusion Refinement, which employs frequency-aware conditional priors and multi-source data to progressively enhance fine-scale precipitation structures while preserving coarse-scale meteorological consistency. Experimental results on the publicly available SEVIR dataset demonstrate that WaveC2R achieves state-of-the-art performance in satellite-based radar retrieval, particularly excelling at preserving high-intensity precipitation features and sharply defined meteorological boundaries.

High-resolution 3D tracking with sub-nanosecond timing is required for the detection of elementary particles, such as neutrinos. Conventional detectors, which utilize analog silicon photomultipliers, face challenges in balancing spatial resolution and scalability. To address this issue, a CMOS single-photon avalanche diode (SPAD)-based high-resolution particle detector is being developed. This work presents a study on SPAD layout optimization and a 4x4 SPAD macropixel module, fabricated in 110 nm CIS technology. Measurement results confirm that high-fill-factor designs improve photon detection efficiency without significant noise degradation. Furthermore, event-driven photon mapping and time stamping, enabled by time-to-digital converters and dedicated pixel circuits integrated into the 4x4 SPAD macropixel, were successfully demonstrated. This work is an essential step towards a sensor that detects probabilistic particle interactions and it lays the groundwork for the development of future large-scale SPAD-based particle detectors.

This paper implements deep reinforcement learning (DRL) for spacecraft reorientation control with a single pointing keep-out zone. The Soft Actor-Critic (SAC) algorithm is adopted to handle continuous state and action space. A new state representation is designed to explicitly include a compact representation of the attitude constraint zone. The reward function is formulated to achieve the control objective while enforcing the attitude constraint. A curriculum learning approach is used for the agent training. Simulation results demonstrate the effectiveness of the proposed DRL-based method for spacecraft pointing-constrained attitude control.

This study presents a novel deep learning approach aimed at enhancing stochastic Gust Load Alleviation (GLA) specifically for compliant wings. The approach incorporates the concept of smooth wing camber variation, where the camber of the wing's chord is actively adjusted during flight using a control signal to achieve the desired aerodynamic loading. The proposed method employs a deep learning-based model predictive controller designed for probability density shaping. This controller effectively solves the probability density evolution equation through a custom Physics-Informed Neural Network (PINN) and utilizes Automatic Differentiation for Model Predictive Control (MPC) optimization. Comprehensive numerical simulations were conducted on a compliant wing (CW) model, evaluating performance of the proposed approach against stochastic gust profiles. The evaluation involved stochastic aerodynamic loads generated from Band-Limited White Noise (BLWN) and Dryden gust models. The evaluation were conducted for two distinct Compliant Chord Fractions (CCF). The results demonstrate the effectiveness of the proposed probability density shaping model predictive control in alleviating stochastic gust load and reducing wing tip deflection.

Many neural networks for point clouds are, by design, invariant to the symmetries of this datatype: permutations and rigid motions. The purpose of this paper is to examine whether such networks preserve natural symmetry aware distances on the point cloud spaces, through the notion of bi-Lipschitz equivalence. This inquiry is motivated by recent work in the Equivariant learning literature which highlights the advantages of bi-Lipschitz models in other scenarios. We consider two symmetry aware metrics on point clouds: (a) The Procrustes Matching (PM) metric and (b) Hard Gromov Wasserstien distances. We show that these two distances themselves are not bi-Lipschitz equivalent, and as a corollary deduce that popular invariant networks for point clouds are not bi-Lipschitz with respect to the PM metric. We then show how these networks can be modified so that they do obtain bi-Lipschitz guarantees. Finally, we provide initial experiments showing the advantage of the proposed bi-Lipschitz model over standard invariant models, for the tasks of finding correspondences between 3D point clouds.

Reducing energy consumption and carbon emissions in data centers by enabling real-time temperature prediction is critical for sustainability and operational efficiency. Achieving this requires accurate modeling of the 3D temperature field to capture airflow dynamics and thermal interactions under varying operating conditions. Traditional thermal CFD solvers, while accurate, are computationally expensive and require expert-crafted meshes and boundary conditions, making them impractical for real-time use. To address these limitations, we develop a vision-based surrogate modeling framework that operates directly on a 3D voxelized representation of the data center, incorporating server workloads, fan speeds, and HVAC temperature set points. We evaluate multiple architectures, including 3D CNN U-Net variants, a 3D Fourier Neural Operator, and 3D vision transformers, to map these thermal inputs to high-fidelity heat maps. Our results show that the surrogate models generalize across data center configurations and significantly speed up computations (20,000x), from hundreds of milliseconds to hours. This fast and accurate estimation of hot spots and temperature distribution enables real-time cooling control and workload redistribution, leading to substantial energy savings (7%) and reduced carbon footprint.

Digital Twin (DT) technologies are transforming manufacturing by enabling real-time prediction, monitoring, and control of complex processes. Yet, applying DT to deformation-based metal forming remains challenging because of the strongly coupled spatial-temporal behavior and the nonlinear relationship between toolpath and material response. For instance, sheet-metal forming by the English wheel, a highly flexible but artisan-dependent process, still lacks digital counterparts that can autonomously plan and adapt forming strategies. This study presents an adaptive DT framework that integrates Proper Orthogonal Decomposition (POD) for physics-aware dimensionality reduction with a Koopman operator for representing nonlinear system in a linear lifted space for the real-time decision-making via model predictive control (MPC). To accommodate evolving process conditions or material states, an online Recursive Least Squares (RLS) algorithm is introduced to update the operator coefficients in real time, enabling continuous adaptation of the DT model as new deformation data become available. The proposed framework is experimentally demonstrated on a robotic English Wheel sheet metal forming system, where deformation fields are measured and modeled under varying toolpaths. Results show that the adaptive DT is capable of controlling the forming process to achieve the given target shape by effectively capturing non-stationary process behaviors. Beyond this case study, the proposed framework establishes a generalizable approach for interpretable, adaptive, and computationally-efficient DT of nonlinear manufacturing systems, bridging reduced-order physics representations with data-driven adaptability to support autonomous process control and optimization.

Deep Brain Stimulation (DBS) is a well-established neurosurgical treatment aiming at symptom alleviation in a range of neurological and psychiatric diseases. Computational models of DBS are widely used to investigate the effects of stimulation on neural tissue, to explore stimulation targets and sweetspots, and ultimately, to aid clinicians in the DBS programming by calculating the stimulation parameters. Commonly, DBS is performed bilaterally, i.e. with one lead in each brain hemisphere, where computational models are solved independently for one lead at a time. This paper treats scenarios where multiple DBS leads are implanted in close proximity to one another, resulting in interacting electrical fields and, therefore, potentially overlapping stimulation spreads. In particular, a global dual-lead model is compared to approximations derived from single-lead approaches in a cohort of twelve multiple sclerosis (MS) tremor patients. It is concluded that simple superposition of volumes of tissue activated (VTAs) underestimates activation, while superposition of electric fields or activating functions leads to overestimation. It is concluded that given close proximity of DBS leads, the VTA cannot be computed individually as stimulation fields exhibit significant and complex interaction. The approach is extended to modeling two obsessive compulsive disorder patients with medially placed leads, where similar VTA discrepancies as in the MS patient cohort are observed.

Image deblurring is vital in computer vision, aiming to recover sharp images from blurry ones caused by motion or camera shake. While deep learning approaches such as CNNs and Vision Transformers (ViTs) have advanced this field, they often struggle with complex or high-resolution blur and computational demands. We propose a new dual-domain architecture that unifies Vision Transformers with a frequency-domain FFT-ReLU module, explicitly bridging spatial attention modeling and frequency sparsity. In this structure, the ViT backbone captures local and global dependencies, while the FFT-ReLU component enforces frequency-domain sparsity to suppress blur-related artifacts and preserve fine details. Extensive experiments on benchmark datasets demonstrate that this architecture achieves superior PSNR, SSIM, and perceptual quality compared to state-of-the-art models. Both quantitative metrics, qualitative comparisons, and human preference evaluations confirm its effectiveness, establishing a practical and generalizable paradigm for real-world image restoration.

We consider the problem of estimating the parameters of a vehicle dynamics model for predictive control in driving applications. Instead of solely using the instantaneous parameters estimated from the vehicle signals, we combine this with cameras and update a probabilistic map with parameter estimates and semantic information using Bayesian moment matching. Key to this approach is the map representation, which is constructed with conjugate priors to the measurement likelihoods and defined in the same path coordinates as the vehicle controller, such that the map can be externalized to provide a local representation of the parameter likelihoods that vary in space. The result is a spatial map of vehicle parameters adapted online to enhance the driving control system. We provide theoretical guarantees on the smoothness of relevant parameter likelihood statistics as a function of space, which is critical for their use in predictive control.

In this work, we address the challenge of generalizable audio deepfake detection (ADD) across diverse speech synthesis paradigms-including conventional text-to-speech (TTS) systems and modern diffusion or flow-matching (FM) based generators. Prior work has mostly targeted individual synthesis families and often fails to generalize across paradigms due to overfitting to generation-specific artifacts. We hypothesize that synthetic speech, irrespective of its generative origin, leaves behind shared structural distortions in the embedding space that can be aligned through geometry-aware modeling. To this end, we propose RHYME, a unified detection framework that fuses utterance-level embeddings from diverse pretrained speech encoders using non-Euclidean projections. RHYME maps representations into hyperbolic and spherical manifolds-where hyperbolic geometry excels at modeling hierarchical generator families, and spherical projections capture angular, energy-invariant cues such as periodic vocoder artifacts. The fused representation is obtained via Riemannian barycentric averaging, enabling synthesis-invariant alignment. RHYME outperforms individual PTMs and homogeneous fusion baselines, achieving top performance and setting new state-of-the-art in cross-paradigm ADD.

In this work, we address the problem of finegrained traceback of emotional and manipulation characteristics from synthetically manipulated speech. We hypothesize that combining semantic-prosodic cues captured by Speech Foundation Models (SFMs) with fine-grained spectral dynamics from auditory representations can enable more precise tracing of both emotion and manipulation source. To validate this hypothesis, we introduce MiCuNet, a novel multitask framework for fine-grained tracing of emotional and manipulation attributes in synthetically generated speech. Our approach integrates SFM embeddings with spectrogram-based auditory features through a mixed-curvature projection mechanism that spans Hyperbolic, Euclidean, and Spherical spaces guided by a learnable temporal gating mechanism. Our proposed method adopts a multitask learning setup to simultaneously predict original emotions, manipulated emotions, and manipulation sources on the EmoFake dataset (EFD) across both English and Chinese subsets. MiCuNet yields consistent improvements, consistently surpassing conventional fusion strategies. To the best of our knowledge, this work presents the first study to explore a curvature-adaptive framework specifically tailored for multitask tracking in synthetic speech.

Electrification is contributing to substantial growth in U.S. commercial and industrial loads, but the cost and Scope 2 carbon emission implications of this load growth are opaque for both power consumers and utilities. This work describes a unique spatiotemporally resolved data set of U.S. electricity costs and emissions and applies time series approximation methods to quantify the alignment of electricity cost and emission incentives for large commercial and industrial consumers. We present a comprehensive spatiotemporal dataset of U.S. price-based demand response (i.e., tariff) and incentive-based demand response programs, enabling direct comparison to previously published marginal emission factor, average emission factor, and day-ahead market prices. We resolved the structural incompatibility and fragmentation of these datasets by developing time series approximations of discrete data and unifying geospatially heterogeneous datasets. Analysis of these datasets reveals significant spatial and temporal heterogeneity in cost and carbon emissions incentives for demand-side energy flexibility, underscoring the importance of site selection as a key factor influencing power costs and Scope 2 emissions. Analysis also reveals broad misalignment of economic and emissions incentives under existing electricity tariff structures, meaning tariffs are incentivizing consumption of more carbon-intensive electricity, and highlighting potential barriers to electrification delivering carbon savings.

For systems whose states implicate sensitive information, their privacy is of great concern. While notions like differential privacy have been successfully introduced to dynamical systems, it is still unclear how a system's privacy can be properly protected when facing the challenging yet frequently-encountered scenario where an adversary possesses prior knowledge, e.g., the steady state, of the system. This paper presents a new systematic approach to protect the privacy of a discrete-time linear time-invariant system against adversaries knowledgeable of the system's prior information. We employ a tailored pointwise maximal leakage (PML) privacy criterion. PML characterizes the worst-case privacy performance, which is sharply different from that of the better-known mutual-information privacy. We derive necessary and sufficient conditions for PML privacy and construct tractable design procedures. Furthermore, our analysis leads to insight into how PML privacy, differential privacy, and mutual-information privacy are related. We then revisit Kalman filters from the perspective of PML privacy and derive a lower bound on the steady-state estimation-error covariance in terms of the PML parameters. Finally, the derived results are illustrated in a case study of privacy protection for distributed sensing in smart buildings.

We propose a joint estimation method for the Direction-of-Arrival (DoA) and the Noise Covariance Matrix (NCM) tailored for beamforming applications. Building upon an existing NCM framework, our approach simplifies the estimation procedure by deriving an quasi-linear solution, instead of the traditional exhaustive search. Additionally, we introduce a novel DoA estimation technique that operates across all frequency bins, improving robustness in reverberant environments. Simulation results demonstrate that our method outperforms classical techniques, such as MUSIC, in mid- to high-angle scenarios, achieving lower angular errors and superior signal enhancement through beamforming. The proposed framework was also fared against other techniques for signal enhancement, having better noise rejection and interference canceling capabilities. These improvements are validated using both theoretical and empirical performance metrics.

Post-disaster situations pose unique navigation challenges. One of those challenges is the unstructured nature of the environment, which makes it hard to layout paths for rescue vehicles. We propose the use of Uncrewed Aerial Vehicle (UAV) in such scenario to perform reconnaissance across the environment. To accomplish this, we propose an optimization-based approach to plan a path for the UAV at optimal height where the sensors of the UAV can cover the most area and collect data with minimum uncertainty.

Current AI systems excel at pattern recognition but fail at causal reasoning. We argue this is not an engineering limitation but reveals something fundamental about the nature of understanding itself. We propose that causal cognition requires a specific physical architecture: stochastic, coupled oscillators with whole-system coordination. To test this, we analyzed high-density EEG (64 channels, 10 subjects, 500 plus trials) from a P300 target recognition task. We computed the Kuramoto Order Parameter (R) to measure global phase synchronization and compared it to standard voltage (ERP) and coherence (ITC) metrics. Four findings establish the framework. Phase and voltage are globally independent (r of 0.048) yet strongly trial-coupled (r of 0.590), proving R captures hidden cognitive structure. Voltage precedes phase by 293 ms, revealing sequential computation. Frequency decomposition shows Theta (169 ms), Alpha (286 ms), and Beta (777 ms) cascade. Our metric is distinct from standard ITC (r of 0.155). Then we tested the standard assumption. Conventional artifact rejection removes eye movements, muscle activity, and autonomic signals before analysis. The standard model assumes cognition is neural activity plus noise. We ran the identical analysis with and without rejection. Removing artifacts reduced the trial-level correlation threefold (from 0.590 to 0.195). Target discrimination reversed sign (from positive 0.6 percent to negative 0.4 percent). What we discard as noise is 70 percent of the signal. This falsifies the standard model. Cognition is not isolated neural computation. It is whole-body phase synchronization spanning neural, muscular, and autonomic systems. For AI, the implications are direct: embodied sensorimotor integration is not optional. It is the substrate that makes understanding possible.

Safe planning of an autonomous agent in interactive environments – such as the control of a self-driving vehicle among pedestrians – poses a major challenge as the behavior of the environment is unknown and reactive to the behavior of the autonomous agent. This coupling gives rise to interaction-driven distribution shifts where the autonomous agent's control policy may change the environment's behavior, thereby invalidating safety guarantees in existing work. Indeed, recent works have used conformal prediction (CP) to generate distribution-free safety guarantees using observed data of the environment. However, CP's assumption on data exchangeability is violated in interactive settings due to a circular dependency where a control policy update changes the environment's behavior, and vice versa. To address this gap, we propose an iterative framework that robustly maintains safety guarantees across policy updates by quantifying the potential impact of a planned policy update on the environment's behavior. We realize this via adversarially robust CP where we perform a regular CP step in each episode using observed data under the current policy, but then transfer safety guarantees across policy updates by analytically adjusting the CP result to account for distribution shifts. This adjustment is performed based on a policy-to-trajectory sensitivity analysis, resulting in a safe, episodic open-loop planner. We further conduct a contraction analysis of the system providing conditions under which both the CP results and the policy updates are guaranteed to converge. We empirically demonstrate these safety and convergence guarantees on a two-dimensional car-pedestrian and a high-dimensional quadcopter case study. To the best of our knowledge, these are the first results that provide valid safety guarantees in such interactive settings.

Hidden Markov Models (HMMs) are fundamental for modeling sequential data, yet learning their parameters from observations remains challenging. Classical methods like the Baum-Welch algorithm are computationally intensive and prone to local optima, while modern spectral algorithms offer provable guarantees but may produce probability outputs outside valid ranges. This work introduces Belief Net, a differentiable filtering framework that learns HMM parameters by formulating the forward filter as a structured neural network and optimizing it with stochastic gradient descent. This architecture recursively updates the belief state, which represents the posterior probability distribution over hidden states based on the observation history. Unlike black-box transformer models, Belief Net's learnable weights are explicitly the logits of the initial distribution, transition matrix, and emission matrix, ensuring full interpretability. The model processes observation sequences using a decoder-only (causal) architecture and is trained end-to-end with standard autoregressive next-observation prediction loss. On synthetic HMM data, Belief Net achieves faster convergence than Baum-Welch while successfully recovering parameters in both undercomplete and overcomplete settings, whereas spectral methods prove ineffective in the latter. Comparisons with transformer-based models are also presented on real-world language data.