Optimization and Control
Papers
Fresh from arXiv, typeset properly. Highlighted words trade on the market.
All of arXiv
Monday 7 September
We study the time-optimal control of a controllable linear system on a time horizon [0,T], focusing on the asymptotic switching density for large T. When the system matrix has only real eigenvalues, it is well-known that the number of switches is upper bounded uniformly in T; when it has complex eigenvalues, no such uniform bound exists, and the switching count instead typically grows with T. We characterize this growth for a system matrix with an arbitrary spectrum, allowing simultaneously for real eigenvalues, complex eigenvalues, and a non-trivial Jordan structure. If the dominant mode is complex, the number of switches grows at least linearly in T, with an explicit lower bound expressed via the mean motion problem and the Bohl-Weyl-Wintner formula. We illustrate the theory on a linearized aircraft pitch/altitude model, showing close agreement between the predicted asymptotic switching rate and numerically computed time-optimal controls.
Robotics
CALM: Configuration-Aware Human Intervention Boundaries During Robot Approach
How robot body configuration shapes human intervention during approach remains underexplored. We conducted a within-participants study with 41 participants, measuring final stopping distance, subjective comfort, and exploratory eye-tracking responses across four humanoid arm configurations and two spatial scales. Full forward arm extension increased stopping distance by approximately 31-36 cm relative to arms-down. Spatial scale primarily affected comfort and pupil responses without a detectable stopping-distance shift. We introduce the Configuration-Aware Limit Model (CALM), which translates stopping-distance distributions into configuration-dependent population-coverage boundaries. Estimated boundaries at 80% coverage ranged from 0.88 to 1.47 m. In an illustrative one-dimensional planning analysis, reconfiguration enabled a 1.10 m approach goal that was unreachable with arms remaining fully extended under the same nominal pointwise 20% intervention-probability constraint. These findings support treating body configuration as a planning variable while distinguishing physical safety, behavioral intervention, and subjective cost.
Read the full abstract
Data Structures and Algorithms
Algorithms for Finite Group Epimorphism Testing
The Group Epimorphism Problem (GpEpi) asks, given two finite groups $G_1$ and $G_2$, whether there exists a surjective group homomorphism, or epimorphism, from $G_1$ to $G_2$. When the input groups are given by their multiplication (Cayley) tables, the problem admits a quasipolynomial-time algorithm in general, but little is known about its complexity for structured classes of finite groups. In this paper, we study the computational complexity of GpEpi for several well-studied classes of finite groups. Our main results are polynomial-time epimorphism tests for several classes of groups for which polynomial-time isomorphism testing was previously known: Groups with Abelian normal Hall subgroups with cyclic complement; Groups with (product of) elementary Abelian normal Hall subgroup with elementary Abelian complement; and Groups with some constraints on their Abelian chief factors.
Read the full abstract
Signal Processing
Masking Radar Cognition under Adversarial Surveillance: A Distributional Privacy Framework
In this article, we propose an online electronic counter-countermeasure (ECCM) framework designed to conceal the strategic decision-making processes of a cognitive radar (CR) operating under adversarial surveillance. We model the CR under two distinct decision paradigms: a static constrained utility-maximizing behavior and a dynamic expected utility-maximizing behavior. The radar's utility function is modeled via a von Mises–Fisher (vMF) distribution, with the distributional parameter constituting the private information to be protected from adversarial inference. We adopt a distribution privacy framework to conceal this private information and provide formal distribution privacy guarantees for cognition masking. In this work, we develop cognition-hiding algorithms for both static constrained utility maximization (WDPCH-SU), and dynamic expected utility maximization (WDPCH-DU). Through rigorous mathematical analysis, we show that both WDPCH-SU and WDPCH-DU satisfy $\epsilon$-distribution privacy ($\epsilon$-DistP) against inference-based adversarial attacks and present the privacy–performance trade-off bounds, quantifying utility loss (in static setting) and expected utility deviation (in dynamic setting) as functions of $\epsilon$. Numerical results show that WDPCH-SU gives about 15% improvement in utility loss at maximum privacy compared to the existing methodology while WDPCH-DU achieves a greater reduction in adversarial Fisher information without requiring explicit Fisher information constraints, at a moderate, analytically bounded utility deviation. These results are highly promising in many 6G communication scenarios such as network slicing for automated driving and swarm UAV coordination, where it is essential to keep the resource allocation policy robust against privacy attacks.
Read the full abstract
Optimization and Control
A Unified Efficient Gradient-Based Heuristic For Box-Constrained Expectation-Related and Risk-Averse Stochastic Optimization Problems
This paper presents a new algorithm addressing the problem of stochastic optimization where the cost function depends on a vector of uncertain parameters with known statistics. The algorithm is parameterized so as to address various stochastic formulations spanning from Expectation-focused to Value-at-Risk (VaR) as well as Conditional-Value-at-Risk (CVaR)-focused formulations. The algorithm leverages a recently proposed gradient-based Search & Accelerate algorithm which is originally dedicated to deterministic optimization problems. The approach is based on a sequence of warm-started solutions of instances of the problem. These solutions together with a samples of other solutions belonging to the convex hull of the first ones constitute the set of admissible candidates. Among this discrete set of candidates, the optimal solution is selected with regards to a sample-based approximation of the targeted criterion. The relevance of the algorithm and its efficiency are discussed and shown using a tailored illustrative example.
Read the full abstract
Mesoscale and Nanoscale Physics
Charge-spin conversion in altermagnets: electronic anisotropy versus magnon-mediated spin drag
Altermagnets acquire a momentum-dependent spin splitting without net magnetization or spin-orbit coupling. This raises the question of whether the transport anisotropy originates in the electrons or the magnons, and whether the magnon-mediated channel distinguishes an altermagnet from a conventional antiferromagnet. We compute the spin-resolved conductivity of a metallic ferromagnet, antiferromagnet, and altermagnet within a common model, treating the spin-diagonal channel with the magnon self-energy and the spin-flip channel to leading order in the electron-magnon coupling. The anisotropy of the charge-spin conversion is dominated by the electronic spin splitting, the magnon branches contributing only marginally. The magnon-mediated spin-flip channel is even under $C_{4}$, so it carries no signature at first order in the anisotropy and cannot on its own distinguish an altermagnet from a conventional antiferromagnet. It does, however, respond linearly to a sublattice-selective perturbation that lowers the bulk $B_{1g}$ symmetry. The resulting susceptibility vanishes identically in the antiferromagnet, is small in the ferromagnet, and is finite in the altermagnet, where the electronic anisotropy ties spin and sublattice together, so that among magnets with no net magnetization it is unique to altermagnetic order. Being even in the magnetic domain index, this response can be measured without preparing a single magnetic-domain sample.
Read the full abstract
Mesoscale and Nanoscale Physics
Layer-Dependent Phonons, Excitons, and Magneto-Optical Phenomena in CrSBr: A Mini Review
Two-dimensional layered magnetic materials offer a versatile platform for exploring low-dimensional magnetism and coupled many-body interactions in these materials. Chromium sulfur bromide (CrSBr) is a promising candidate for advanced spintronic and optoelectronic applications because of its intrinsic air stability, semiconducting nature, strong in-plane anisotropy, and A-type antiferromagnetic ordering. This review summarizes recent advances in the understanding of the layer-dependent vibrational and excitonic properties of CrSBr, as well as its magneto-optical response, from bulk crystals to the monolayer limit. We examined its crystal structure, magnetic anisotropy, and interlayer spin reorientation, followed by insights into vibrational dynamics and spin-phonon coupling. Particular emphasis is placed on the excitonic landscape, including magnetic-field-sensitive photoluminescence, localized excitonic states, and the coexistence of Frenkel- and Wannier-Mott excitons in the bandgap. Finally, we discuss the challenges and prospects of harnessing the unique layer-dependent properties of CrSBr in spintronic, magneto-optical, and quantum photonic technologies.
Read the full abstract
High Energy Physics (Phenomenology)
Connecting Regge factorization and Collins-Soper evolution
The relationship between Regge factorization of amplitudes and the Collins-Soper-Sterman factorization of hadronic cross sections is explored. Seeking a unified description of rapidity divergences in high-energy scattering and transverse-momentum-dependent factorization, a connection emerges between the Collins-Soper kernel and the Regge trajectory. This provides a framework for bridging high-energy and low-transverse-momentum resummation and for future extensions of $k_T$-factorization beyond leading-logarithmic accuracy.
Read the full abstract
Differential Geometry
Yau's conjecture for the stacked Clifford tori of Wiygul
We prove Yau's conjecture $\lambda_{1}=2$ for the stacked Clifford tori of Wiygul: for all integers $N\ge2$, $k,\ell\ge1$ and every sufficiently large $m$, every closed embedded minimal surface arising from Wiygul's construction, of genus $k\ell m^{2}(N-1)+1$ in the round three-sphere and resembling $N$ parallel copies of the Clifford torus joined by small catenoidal tunnels, has first Laplace eigenvalue $2$. For $N\ge3$ these surfaces are chains rather than doublings, and the even–odd decomposition on which all previous verifications for gluing constructions rest is not available. The reflection lemma of Choe and Soret reduces the problem to the sector of functions invariant under the symmetry group of the construction, and we show that the lowest nonzero eigenvalue of that sector equals $4+O(m^{-1})$. The value $4$ is the outcome of an exact identity in the limiting weighted graph model: the limiting waist ratios of the construction form the Perron vector of the adjacency operator of the line graph of a path, so that the spectral gap of the path cancels against the total conductance of the tunnels prescribed by the balancing conditions, and what survives is the coefficient of the Jacobi operator of the Clifford torus. The analytic input consists of a conformally invariant channel inequality on a cylinder and of a Poincaré inequality on a periodically perforated torus. The properties of the construction on which the argument rests are isolated in a reduction theorem for closed surfaces decomposed into blocks joined by families of thin channels along the edges of a finite graph, subject to a symmetry assumption and to Poincaré and trace inequalities on the blocks.
Read the full abstract
Classical Analysis and ODEs
Two-parameter variational estimates for averages over tori
One-parameter variational inequalities are well developed, whereas their multi-parameter counterparts remain much less understood. We explore two-parameter variational inequalities for averages over tori in $\mathbb{R}^3$. To capture the underlying two-parameter structure, we introduce a local two-parameter $r$-variation norm that combines rectangular increments with variations along the boundary. The resulting variation operator pointwise dominates the corresponding two-parameter local maximal function and, unlike the maximal function, also captures oscillation across the two parameters. We establish sharp $L^p$–$L^q$ bounds for this variation operator up to endpoints. For comparison, we also obtain sharp $L^p$ bounds up to endpoints for the corresponding local one-parameter variation operator, revealing a genuine difference between the one- and two-parameter boundedness regions. The proof combines square function estimates for two-parameter propagators with local smoothing estimates through mixed-norm interpolation.
Read the full abstract
Materials Science
Quantifying thermal and driven magnon populations with femtosecond noise correlation spectroscopy
Precise knowledge of the total number of magnons, including both coherent and incoherent (e.g. thermal) excitations, is imperative for the advancement of fundamental spin-wave physics and the development of next-generation magnonic devices. In particular, quantifying magnons is key to understanding magnon transport phenomena, the nonlinear regime, or ultrafast magnetization dynamics. Typically, incoherent magnons are accessed by frequency-domain techniques, which lack the temporal resolution required for ultrafast processes, while ultrafast time-domain methods are generally sensitive only to the coherent dynamics. In this work, we demonstrate that femtosecond noise correlation spectroscopy enables a fully quantitative, time-domain measurement of both thermal and coherently excited magnon modes in bismuth-substituted yttrium iron garnet driven by a free-running microwave. We model the experimental data and extract the magnon number by simulating the magnon band structure of the sample, the magneto-optical response function, and the optical spot size used in the experiment. Our analysis establishes a connection between magnon mode calculations and experimentally accessible magnetic properties and fiducially reproduces the waveform and amplitude of the magneto-optical correlation signal for different experimental conditions. These results open a new pathway towards the optical tomography of magnon modes in non-linear or non-equilibrium conditions and can be readily extended to study ultrafast incoherent dynamics in other condensed matter systems.
Read the full abstract
Computer Vision
Unified Vision-Centric Pedestrian Crossing Action Prediction via Adaptive Patch Projection and Proactive Spatial Rectification
Vision cues are available and informative for pedestrian action prediction, but obtaining stable target-centric representations from video frames remains challenging without frame-level external perception cues. Thus, most methods rely on additional perception modules or multi-source information fusion, leaving the reliability of vision-centric setting an open question. To this end, we propose ViCross, a vision-centric pedestrian crossing action prediction framework powered by multimodal large language models, which maintains target-centric reasoning from video frames without additional perception modules beyond first-frame target initialization. While multimodal large language models exhibit strong visual understanding, applying them directly to vision-centric action prediction faces two challenges. First, accurately perceiving target pedestrians often requires high resolution inputs and dense visual tokenization, making full-frame encoding computationally prohibitive. ViCross tackles this with Variable Resolution Patch Mapping module for efficient token allocation while preserving key pedestrian details. Second, missing spatiotemporal priors hinder consistent cross frame reasoning. ViCross mitigates this with a Spatial Constraint Enhancement Strategy that captures past motion, future locations, and action semantics for training-time proactive spatial rectification. Extensive experiments show that ViCross delivers clear gains in vision-centric prediction settings and is competitive with multi-source fusion approaches in several settings. Code is available at https://github.com/2tianyao1/ViCross.git.
Read the full abstract
Number Theory
On Dynamical Mordell-Lang Conjecture for Split Map
In this article, we prove the dynamical Mordell-Lang conjecture for product of endomorphisms on curves defined over complex number field. Moreover, we proved that there is some effectively computable on the the common difference of the arithmetic progressions.
Neural and Evolutionary Computing
Photonic reservoir computing with dimensionally compressed readout
This work addresses a hardware constraint in reservoir computing: the limited size of the readout layer imposed by systems with a physical readout. We investigate a strategy to accommodate this constraint based on random projection, which compresses high-dimensional reservoir states into a lower-dimensional subspace while preserving key properties of the source space and information- processing capabilities. To evaluate this approach, we compare a small, standalone time delay reservoir against a larger configuration whose output is projected down to match the same restricted readout dimension. Using task-independent metrics, we demonstrate that the distribution of information-processing capacities may differ between the two configurations, even at identical readout sizes. Furthermore, we perform a comprehensive hyperparameter scan to assess how both systems behave under varying physical regimes. Finally, we benchmark this approach on the standard NARMA10 task, showing that the random projection framework can yield superior performance compared to a standalone constrained reservoir, within specific compression range. These results provide a scalable pathway to bypass physical readout bottlenecks in hardware-based reservoir computing.
Read the full abstract
High Energy Astrophysical Phenomena
Oblique Collision of a Relativistic Cold Shell with an Ideal Reflecting Wall
Relativistic flows are common in astrophysics and often form shocks when different parts of the flow collide at relativistic relative velocities. Such collisions are often oblique, forming two shocks whose shocked fluids are separated by a contact discontinuity, which is treated here as an ideal reflecting “wall” where the flow on either side is modeled separately. The latter is modeled in the lab frame $S$ as a uniform cold planar shell propagating into vacuum at velocity $v_1=\beta_1c$ normal to its vacuum interface, colliding with the wall at an incidence angle $\alpha_1$. The collision point $P$ moves along the wall at a velocity $v_p=v_{1}/\sin\alpha_1$, and a boost along the wall at $v_p$ leads to a steady-state frame $S'$ where this problem is highly simplified. However, a “super-luminal” regime exists where $v_p>c\Leftrightarrow\tan\alpha_1<\Gamma_{1}\beta_{1}=(1-\beta_{1}^2)^{-1/2}\beta_1$ and no steady-state frame $S'$ exists. It corresponds to only very small $\alpha_1$ in the Newtonian regime, but nearly all $\alpha_1$ in the relativistic regime. We solve this problem \textit{fully analytically} using integral conservation laws, in the attachmrnt region where point $P$ is attached to the wall. This region of parameter space is bound at high $\alpha_1$ by the detachment line, which coincides with the sonic line for a cold initial shell. A weak-shock solution exist in all this region, while a strong-shock solution exists only in the sub-luminal attachment region – between the luminal line and the detachment/sonic line where the two solutions coincide and beyond which point $P$ detaches from the wall and shocked fluid spills into the vacuum.
Read the full abstract
Quantum Physics
Photonic Time-Delayed Quantum Extreme Learning Machine
We propose and numerically simulate a photonic quantum extreme learning machine (QELM) based on non-linear interactions in a time-delay non-linear interferometer with photon-number-resolving detection. We theoretically demonstrate the QELM's computational capability by binary classification of the non-linear two moons benchmark. Analysis of the learned output weights and alternative quantum readouts shows that compact feature sets based on photon-number correlations preserve classification accuracy while potentially reducing measurement complexity. We further show that the QELM maintains its classification performance under the effect of losses by naturally redistributing the learned weights across a larger set of available features. These results highlight the potential of time-delay photonic implementation as an experimentally accessible QELM platform.
Read the full abstract
Computer Vision
Self-Supervised Multi-View 3D Gaze Target Estimation via Probabilistic Ray Marching
We present a self-supervised approach, Self-MVGTE, for estimating 3D gaze targets from multiple camera views. Unlike existing methods that independently estimate 2D gaze targets per camera view, Self-MVGTE predicts gaze targets directly in 3D space for the first time. Moreover, it does not require any ground-truth annotations from the target scene and uses only the multi-view input images from a calibrated camera setup, pseudo 2D gaze target labels from a monocular gaze target estimation model, and 3D gaze vectors from a monocular 3D gaze estimation model. A key challenge is that these pseudo labels are inherently noisy and multi-view inconsistent. To address this, we propose a probabilistic ray marching framework, which models the uncertainty of these pseudo labels and exploits 3D gaze vectors as geometric priors. Specifically, these gaze vectors are first integrated into the monocular gaze target estimation model to improve its generalization to unseen scenes, producing higher-quality pseudo labels. Then, for 3D gaze target estimation, we construct a 3D gaze cone by casting a bundle of rays from the eye position around the gaze vector to strictly constrain the solution space. Within this cone, we propose a depth-guided feature sampling strategy using off-the-shelf DINOv2 and Depth-Anything-3 models, and estimate a spatial likelihood distribution of the gaze target. Finally, we convert the pseudo gaze target labels into a target distribution and softly optimize the network. Extensive experiments on the MVGT dataset show that Self-MVGTE achieves state-of-the-art performance, surpassing existing fully-supervised baselines.
Read the full abstract
Computer Vision
RelightFormer: Feed-forward Generative Transformer for Multiview Object Relighting
Image relighting is traditionally tackled via complex inverse rendering pipelines, which suffer from ill-posed optimization, or single-image generative models that ignore crucial multi-view cues necessary for understanding 3D geometry and material interactions. To address these limitations, we introduce a feed-forward generative Transformer for direct single- and multi-view image relighting that entirely bypasses explicit intrinsic property estimation. Adapted from a video foundation model, our architecture features a latent illumination module that dynamically injects target environment maps into spatial features via cross-attention. Furthermore, we employ permutation-invariant positional encodings to symmetrically process unordered multi-view inputs without sequential bias. To train this robust data-driven model, we construct the massive Laval Objaverse Dataset (LOD), comprising 90K objects and 39K unique illuminations. Extensive experiments demonstrate state-of-the-art visual quality, photorealistic relighting quality, and strong zero-shot generalization across single-view, multi-view, and novel-view relighting tasks.
Read the full abstract
General Relativity and Quantum Cosmology
Binary Black Holes in Einstein-Maxwell-Dilaton Theory: Second Post-Newtonian Dynamics from Effective Field Theory
The detection of gravitational waves from compact binary coalescences provides a powerful opportunity to test general relativity in the strong-field regime and to search for signatures of alternative theories of gravity. In this work, we consider Einstein-Maxwell-Dilaton (EMd) theory, in which black holes can carry both electric and scalar (dilatonic) charges. We employ the effective field theory approach, together with a temporal Kaluza-Klein decomposition of the metric in terms of non-relativistic gravitational fields, to derive the conservative two-body Lagrangian for charged black-hole binaries in EMd theory through second post-Newtonian (PN) order. Our calculation extends the previously known conservative dynamics at 1PN order and includes the gravitational, electromagnetic, and dilaton interactions at 2PN order. We verify the result in the appropriate Einstein-Maxwell, scalar-tensor, and general relativistic limits, and perform an independent test-body-limit check of the static 2PN sector. These results provide the conservative dynamics needed for developing higher-accuracy waveform models and testing EMd theory with gravitational-wave observations.
Read the full abstract
High Energy Astrophysical Phenomena
Timing Gravity with Pulsars in the Strong Field
We propose a novel approach for the timing of pulsars orbiting a supermassive black hole, which implements the fully relativistic calculations of the photon travel time into a robust timing model. We generate realistic mock catalogues of pulsar times-of-arrival for several putative pulsars on tight orbits around the Galactic Center supermassive black hole, Sagittarius A* (Sgr A*). Then, we perform a proof-of-concept sensitivity analysis to forecast the accuracy that future observational facilities, like the Squared Kilometer Array, will achieve in the characterization of the parameters of our timing model. Our analysis shows how the observation of pulsars at the Galactic Center will open an incredibly promising avenue for the characterization of the physical properties of Sgr A*, which can improve by at least three orders of magnitude the current constraints on the black hole's mass achieved with the S-stars and event-horizon scale observations.
Read the full abstract
Analysis of PDEs
Threshold phenomena for least energy solutions of a doubly critical Neumann problem with a critical absorption term
We investigate the existence and nonexistence of least energy solutions for a doubly critical Neumann problem with a critical absorption term. We consider a positive function $u$ defined on a smooth bounded domain $\Omega\subset\mathbb{R}^n$ with $n\ge 5$, satisfying $-\Delta u + \lambda u = u^{2^* - 1} -\alpha u^{2^\sharp-1}$ inside $\Omega$, with Neumann boundary condition $\nabla u\cdot \nu = u^{2^\sharp -1}$ on $\partial\Omega$, where $\lambda>0$, $\alpha\ge0$, $2^*=\frac{2n}{n-2}$ is the critical Sobolev exponent and $2^\sharp=\frac{2(n-1)}{n-2}$ denotes the critical trace exponent. The simultaneous presence of the interior and boundary critical exponents together with this critical absorption term creates a new interaction between competing concentration mechanisms. In particular, the absorption term has the same critical trace exponent as the boundary nonlinearity, but with the opposite sign and acting in the interior of $\Omega$. Consequently, it competes with the boundary mean curvature correction at the same asymptotic order in the energy expansion, leading to a sharp threshold phenomenon. Combining several analytic techniques and geometric tools, we prove the existence of a threshold value $\alpha_{0}=\alpha_{0}(\lambda,\Omega)\in(0,+\infty)$ such that the problem admits a least energy solution if $\alpha<\alpha_{0}$, and no least energy solution if $\alpha>\alpha_{0}$. Moreover, the problem also has a least energy solution at $\alpha =\alpha_0$ provided $\alpha_0> C(n)\max_{\partial \Omega} H$, where $C(n)$ is a positive constant depending only on $n$ and $H$ is the mean curvature on $\partial\Omega$.
Read the full abstract
Computer Vision
Multi-label versus multi-class classification of blood cells and their aggregates in microfluidic channels
Deformability cytometry (DC) is a type of imaging flow cytometry, which uses a camera-equipped device to measure cellular stiffness in addition to other cellular properties at high throughput. Cellular properties such as area and elongation can identify cell types, but this requires prior knowledge of distinguishing properties and cannot be applied to clinically important cell aggregates. Using DC data, we evaluated conventional multi-class (MC) classification and introduced a multi-label (ML) approach for identifying blood cells and their aggregates. In particular, an ML classifier can simultaneously assign multiple cell-type labels to a single imaged event. We show that, unlike MC classification, ML classification can identify cell aggregates not represented in the training data. It also avoids the need for exhaustive, strictly defined aggregate labels, thereby simplifying and speeding up annotation. Since automated blood analyzers do not reliably analyze cell aggregates, our approach may help address this clinical gap.
Read the full abstract
Artificial Intelligence
RAFM-SER++: A Lightweight Multimodal Emotion Recognition Framework for Real-Time Behavioral Monitoring in Surveillance Systems
Recent multimodal Speech Emotion Recognition (SER) systems achieve high accuracy through interaction-heavy cross-modal transformers, but their computational cost limits deployment in latency-sensitive and resource-constrained surveillance systems. To address this challenge, we propose RAFM_SER++, a lightweight multimodal SER framework featuring an asymmetric Residual Attention Fusion Mechanism (RAFM). Rather than relying on computationally expensive bidirectional interactions, RAFM injects affective speech cues into semantic text representations through a one-directional residual attention pathway. Combined with a BYOL-inspired cross-modal alignment objective and attention-guided pooling, the proposed framework improves multimodal representation learning while maintaining low computational overhead. Experiments on the IEMOCAP and ESD benchmarks demonstrate that RAFM_SER++ consistently outperforms the HuBERT-Base baseline and achieves a superior accuracy-efficiency trade-off compared with the state-of-the-art MemoCMT. Specifically, RAFM_SER++ reduces trainable parameters by more than 60%, achieves faster inference (79.60 it/s), and attains BACC scores of 81.10% on IEMOCAP and 95.39% on ESD. These results indicate that lightweight asymmetric multimodal fusion is an effective alternative to interaction-heavy cross-modal transformers for real-time surveillance applications.
Read the full abstract
Machine Learning
Impact of canny edge detection preprocessing on performance of machine learning models for Parkinson's disease classification
This study investigates the classification of individuals as healthy or at risk of Parkinson's disease using machine learning (ML) models, focusing on the impact of dataset size and preprocessing techniques on model performance. Four datasets are created from an original dataset: DS_0, (normal dataset), DS_1 (DS_O subjected to Canny edge detection and Hessian filtering), DS_2 (augmented DS_0), and DS_3 (augmented DS_1). We evaluate a range of ML models-Logistic Regression (LR), Decision Tree (DT), Random Forest (RF), Gradient Boosting (GB), XGBoost (XBG), Naive Bayes (NB), Support Vector Machine (SVM), and AdaBoost (AdB)-on these datasets, analyzing prediction accuracy, model size, and prediction latency. The results show that while larger datasets lead to increased model memory footprints and prediction latencies, the Canny edge detection preprocessing supplemented by Hessian filtering (used in DS_1 and DS_3) degrades the performance of most models. In our experiment, we observe that Random Forest (RF) maintains a stable memory footprint of 61 KB across all datasets, while models like KNN and SVM show significant increases in memory usage, from 5.7-7 KB on DS_0 to 102-220 KB on DS_2, and similar increases in prediction time. Logistic Regression, Decision Tree, and Naive Bayes show stable memory footprints and fast prediction times across all datasets. XGBoost's prediction time increases from 180-200 ms on DS_0 to 700-3000 ms on DS_2 (truncated)
Read the full abstract
Classical Analysis and ODEs
Admissible characterization in delay equations and robustness of exponential dichotomies against small-delay perturbations
Robustness of exponential dichotomies against small-delay perturbations presents a fundamental obstacle: the lack of an effective admissible characterization in nonautonomous delay equations. Considerable efforts have been devoted to obtaining such a characterization for differential equations in Banach spaces. However, even unlike ordinary differential equations, the variation of constants formula for delay equations requires extending the phase space to a space of discontinuous functions, and the dependence on the past states leads to another challenge in establishing exponential dichotomy via admissibility, i.e., constructing a suitable admissible pair and then estimating the growth/decay rates along the unstable/stable subspaces via the admissibility property. In this paper, we provide an admissible characterization based on two pairs of Banach spaces that yields explicit dichotomy exponents. Through this characterization and an operator perturbation method, we prove that small-delay perturbations preserve exponential dichotomies.
Read the full abstract
Machine Learning
Think Wider: Mitigating Latent Rank Collapse in Implicit Chain-of-Thought Reasoning
Chain-of-thought (CoT) reasoning improves the reasoning ability of large language models by introducing intermediate computation, but explicit rationales increase decoding length, latency, and context cost. Implicit CoT offers a more efficient alternative by moving intermediate reasoning into continuous latent states. However, latent reasoning can be unstable: successive latent states may become overly similar and collapse toward a shared dominant direction, reducing the diversity of the reasoning trajectory. In this work, we identify $\textit{latent rank collapse}$ and propose $\textbf{WIDER}$, a lightweight spectral regularizer for implicit CoT. During training, WIDER estimates the shared direction of each latent trajectory and penalizes projections onto this direction, encouraging latent states to span a broader representational subspace. The method is plug-and-play and leaves the backbone model, latent schedule, and inference-time decoding procedure unchanged. We further formulate this collapse as a geometric bottleneck in implicit reasoning, casting its mitigation as a training-time regularization problem rather than an inference-time decoding change. Extensive experiments show that WIDER improves matched implicit CoT baselines, while mechanistic analyses reveal higher effective rank, lower dominant-direction energy, and reduced redundancy among latent steps. These results highlight latent subspace utilization as an important factor for efficient continuous reasoning, providing a geometric perspective for analyzing and improving implicit CoT. Code is available at https://github.com/whitesweater/WIDER.
Read the full abstract
Quantum Gases
Spontaneous counterflow in rotating supersolids
Rotating supersolids challenge the conventional notion of superfluid flow: quantized vortices coexist with crystalline order arising from the system's intrinsic density modulation. Drawing on concepts developed in the context of superfluid helium, this seemingly paradoxical state is often described phenomenologically in terms of a dual response, combining superfluid flow with rigid-body-like motion. Here we show that these apparently distinct behaviors emerge from a single microscopic velocity field, globally constrained by irrotationality. By applying a Helmholtz decomposition, we disentangle a divergence-free incompressible component, which carries vortex circulation, from a curl-free compressible component generated by the intrinsic density modulation. In absence of vortices, the response is entirely encoded in the compressible field: the droplets co-rotate with the external drive, while the interstitial fluid develops an oppositely directed counterflow that preserves vanishing circulation. Above the vortex nucleation threshold, the incompressible component carries the usual quantized circulation, whereas the compressible field counteracts the vortex-induced angular momentum, reducing the total jump with respect to an unmodulated condensate. Our results recast supersolid rotation in terms of a flow-counterflow response, providing a microscopic interpretation of the phenomenological rigid/superfluid partitioning.
Read the full abstract
Earth and Planetary Astrophysics
Long-baseline ATLAS photometry of the seven main-belt asteroid targets of the Emirates Mission to the Asteroid Belt
The Emirates Mission to the Asteroid Belt (EMA) will use the MBR Explorer spacecraft to fly past six main-belt asteroids – (10253) Westerwald, (623) Chimaera, (13294) Rockox, (88055) Ghaf, (23871) Ousha and (59980) Moza – before rendezvousing with, and deploying a lander onto, the extremely red (269) Justitia. We present a homogeneous analysis of a decade of sparse photometry of all seven targets from the Asteroid Terrestrial-impact Last Alert System (ATLAS). For each object we derive absolute magnitudes and phase slopes in the ATLAS c and o bands, the c-o colour on a common phase slope, a rotation period from a Lomb-Scargle periodogram of the phase-corrected light curve, and a light-curve amplitude setting a lower limit on elongation. Five periods are secure – 3.243 h for Rockox, 33.1 h for Justitia, 3.64 h for Westerwald, 14.6 h for Chimaera and 4.36 h for Moza – and agree with independent work; for Ousha aliasing leads us to adopt the published 8.35 h, while Ghaf remains unresolved. Lightcurve inversion gives a unique convex shape and spin state for Rockox and two pole solutions for Chimaera, Moza and Justitia; no model was obtained for Ghaf, Ousha or Westerwald, though Westerwald's spin axis must be retrograde. Combining the c-o colour with an albedo class from the phase slope, we classify the sample as one dark C-type (Chimaera), three probable S-types (Westerwald, Ghaf, Moza), two X-complex objects (Rockox, plausibly M-type, and Ousha, possibly E-type), and the exceptionally red L/D-type Justitia, whose (c-o) = 0.48 is redder than typical Jupiter Trojans. This only partly matches the mission's family-based expectation of a predominantly primitive target set, so we present these as testable predictions. Those of Ghaf and Ousha are least secure – both disagree with published work and have unresolved or aliased periods – and warrant follow-up.
Read the full abstract
Computer Vision
Federated Binary Gating with Server-Side Vision-Language Inference for Surveillance Anomaly Classification
Privacy-sensitive surveillance systems could benefit from large vision-language models (VLMs), but such models typically require centralized access to raw video. In federated learning settings, this challenge is amplified by non-independent and identically distributed (non-IID) client data, which can make direct multiclass anomaly classification unstable, especially for rare categories. We propose a hybrid two-stage architecture that combines a federated binary convolutional neural network (CNN) gate with server-side zero-shot VLM inference. The lightweight LiteCNN3D gate performs local anomaly screening and forwards only flagged videos to Qwen3-VL-8B, which assigns them to four anomaly metaclasses. We evaluate this design on UCF-Crime grouped into five coarse metaclasses and implement the federated stage in a real three-node heterogeneous deployment. In the studied setting, direct federated multiclass training collapses, whereas the proposed decomposition yields a better trade-off between classification quality and raw-video transmission. With fixed-threshold routing, the federated hybrid pipeline preserves nearly the same macro-averaged F1 score (F1-macro) as its centralized CNN+VLM counterpart while reducing the fraction of transmitted videos to 51.4%, although with a lower proxy macro receiver operating characteristic area under the curve (ROC AUC) than the centralized hybrid system. A complementary sensitivity-oriented routing operating point increases macro ROC AUC from 0.673 to 0.692 and reduces the false negative rate from 29.3% to 22.9%, but decreases F1-macro from 0.503 to 0.485 while increasing transmission from 51.4% to 57.9%. These results suggest that federation is better suited to coarse local screening, while routing rules can be adjusted to trade server-side VLM usage for higher anomaly sensitivity.
Read the full abstract
Data Structures and Algorithms
Better Late Than Never: Online Flow Time Scheduling with Online Estimates
In the classical online flow-time scheduling problem on a single machine, jobs arrive over time and must be processed to minimize the total time they spend in the system: for over fifty years, we have known that SRPT is an optimal online algorithm. But this algorithm requires exactness in two different ways: (a) job sizes must be known exactly, and (b) they must be revealed as soon as the job arrives. Recent work relaxed each of these assumptions separately: there are algorithms based on knowing approximate sizes (given when the job arrives), or based on knowing (exact) sizes at some point before the remaining size gets too small. Nonetheless, prior to this work, there was no known approach to relax both assumptions simultaneously. In this work, we consider a model that demands much less: When we process a job, at some point in time between when we complete an $\varepsilon$-fraction and a $(1-\varepsilon)$-fraction of its unknown processing requirement, we are informed that the job is “somewhere in the middle”. Finally, when the job has received its desired amount of processing, we are informed of its completion. No other information is shared about the job. We give an $O(1/\varepsilon^2)$-competitive algorithm for this model. Slightly more generally, we assume that an algorithm receives a $\mu$-approximate estimate of each job's processing time at some time before we complete a $(1-\varepsilon)$-fraction of its processing. Our algorithm is $O(\mu/\varepsilon)$-competitive, and we show that this is asymptotically optimal. It is a surprisingly natural variant of the multilevel feedback algorithm (MLF) and it is parameter-oblivious: it does not need to know $\mu$ or $\varepsilon$ upfront. The core analytical contribution is to robustify the dual-fitting framework for this problem to handle jobs for which we have not yet received estimates.