arXiv subjects

Rui Chen

Publications and source records attributed to Rui Chen.

At least 127 records · Page 7Linked to original sources

arXiv2026

Deep Learning-Based Meat Freshness Detection with Segmentation and OOD-Aware Classification

In this study, we present a meat freshness classification framework from Red-Green-Blue (RGB) images that supports both packaged and unpackaged meat datasets. The system classifies four in-distribution (ID) meat classes and uses an out-of-distribution (OOD)-aware abstention mechanism that flags low-confidence samples as No Result. The pipeline combines U-Net-based segmentation with deep feature classifiers. Segmentation is used as a preprocessing step to isolate the meat region and reduce background, producing more consistent inputs for classification. The segmentation module achieved an Intersection over Union (IoU) of 75% and a Dice coefficient of 82%, producing standardized inputs for the classification stage. For classification, we benchmark five backbones: Residual Network-50 (ResNet-50), Vision Transformer-Base/16 (ViT-B/16), Swin Transformer-Tiny (Swin-T), EfficientNet-B0, and MobileNetV3-Small. We use nested 5x3 cross-validation (CV) for model selection and hyperparameter tuning. On the held-out ID test set, EfficientNet-B0 achieves the highest accuracy (98.10%), followed by ResNet-50 and MobileNetV3-Small (both 97.63%) and Swin-T (97.51%), while ViT-B/16 is lower (94.42%). We additionally evaluate OOD scoring and thresholding using standard OOD metrics and sensitivity analysis over the abstention threshold. Finally, we report on-device latency using TensorFlow Lite (TFLite) on a smartphone, highlighting practical accuracy-latency trade-offs for future deployment.

cs.LG↗

arXiv2026

Oxygen permeability and stability in the entropy-stabilized Co-based Perovskite oxygen permeable membranes

Oxygen transport membranes (OTMs), enabling catalytic reaction and gas separation, support crucial chemical engineering processes and decarbonization technologies, but their applications are hindered by limited oxygen permeation fluxes and inadequate long-term stability during operation. Here, a series of high-entropy perovskite OTMs based on La0.5Sr0.5CoO3 were designed and synthesized by the simple sol-gel method. The impact of varying doping ratios on the structure, surface morphology, oxygen permeability, and stability of these high-entropy OTMs was thoroughly examined. At 950 °C, the optimal composition, La0.25Sr0.25Gd0.2Nd0.2Pr0.1CoO3, achieved oxygen permeation fluxes of 1.62 mL min-1 cm-2 under air/He gradient and 1.46 mL min-1 cm-2 under air/CO2, respectively. Remarkably, all high-entropy OTMs demonstrated stable operation for over 100 h in a pure CO2 environment without a significant decline in performance. This finding paves a new way to enhance the structural and oxygen permeation stability of OTMs, and further promotes the application of OTMs in oxy-fuel combustion technologies aimed at improving CO2 capture and storage efficiency.

cond-mat.mtrl-sci↗

arXiv2026

Adaptive Multi-Objective Tiered Storage Configuration for KV Cache in LLM Service

The memory-for-computation paradigm of KV caching is essential for accelerating large language model (LLM) inference service, but limited GPU high-bandwidth memory (HBM) capacity motivates offloading the KV cache to cheaper external storage tiers. While this expands capacity, it introduces the challenge of dynamically managing heterogeneous storage resources to balance cost, throughput, and latency under varying workloads. We formulate this as a multi-objective optimization problem: identifying the Pareto frontier across these metrics within the storage configuration space. Using a high-fidelity end-to-end simulator, we observe that the objective functions are non-analytic and exhibit complex variable coupling, making the Pareto frontier difficult to approximate analytically. To obtain the frontier, we introduce Kareto, a KV-cache Adaptive REsource managemenT Optimizer. Kareto leverages a diminishing-return-guided pruning method to efficiently navigate the large configuration space and approximate the Pareto frontier. Additionally, it incorporates a fine-grained adaptive tuner that uses eviction policies in tier storage and KV block access patterns for group-specific cache management, improving cache efficiency. Experiments on real-world traces show that Kareto adapts to workload and can identify configurations of better cost efficiency, covering static strategies. Compared to the fixed setup with 1024 GB DRAM, Kareto can improve throughput by up to 9.3%, or reduce latency by up to 58.3%, or lower cost by up to 20.2% under respective optimization objectives.

cs.AR↗

arXiv2026

Design and Characterization of a Dual-DOF Soft Shoulder Exosuit with Volume-Optimized Pneumatic Actuator

Portable pneumatic systems for 2 degree-of-freedom (DOF) soft shoulder exosuits remain underexplored, and face fundamental trade-offs between torque output and dynamic response that are further compounded by the need for multiple actuators to support complex shoulder movement. This work addresses these constraints through a volume-optimized spindle-shaped angled actuator (SSAA) geometry: by reducing actuator volume by 35.7% (357mL vs. 555mL), the SSAA maintains 94.2% of output torque while achieving 35.2% faster dynamic response compared to uniform cylindrical designs. Building on the SSAA, we develop a curved abduction actuator (CAA) based on the SSAA geometry and a horizontal adduction actuator (HAA) based on the pouch motor principle, integrating both into a dual-DOF textile-based shoulder exosuit (390 g). The exosuit delivers multi-modal assistance spanning shoulder abduction, flexion, and horizontal adduction, depending on the actuation. User studies with 10 healthy participants reveal that the exosuit substantially reduces electromyographic (EMG) activity across both shoulder abduction and flexion tasks. For abduction with HAA only, the exosuit achieved up to 59% muscle activity reduction across seven muscles. For flexion, both the single-actuator configuration (HAA only) and the dual-actuator configuration (HAA,+,CAA) reduced EMG activity by up to 63.7% compared to no assistance. However, the incremental benefit of adding the CAA to existing HAA support was limited in healthy users during flexion, with statistically significant additional reductions observed only in pectoralis major. These experimental findings characterize actuator contributions in healthy users and provide design guidance for multi-DOF exosuit systems.

cs.RO↗

arXiv2026

Large-Time Asymptotics for Heat and Fractional Heat Equations on the Lattice and General Finite Subgraphs

In this paper, we study large-time asymptotics for heat and fractional heat equations in two discrete settings: the full lattice $\mathbb Z^d$ and finite connected subgraphs with Dirichlet boundary condition. These results provide a unified discrete theory of long-time asymptotics for local and nonlocal diffusions. For $d\ge1$ and $s\in(0,1]$, we consider on $\mathbb Z^d$ the Cauchy problem \[ \partial_t u+(-Δ)^s u=0,\qquad u(0)=u_0\in \ell^1(\mathbb Z^d), \] and derive a precise first-order asymptotic expansion toward the lattice fractional heat kernel $G_t^{(s)}$. The main technical input is a pair of sharp translation-increment bounds for $G_t^{(s)}$: a pointwise estimate and an $\ell^1$-estimate. As consequences, under finite first moment we obtain the optimal decay rate $t^{-1/(2s)}$ in $\ell^p$-asymptotics ($1\le p\le\infty$), and we prove sharpness by explicit shifted-kernel examples. Without moment assumptions, we still establish convergence in the full $\ell^1$-class, and we show that no universal quantitative rate can hold in general. We also analyze fractional Dirichlet diffusion on finite connected subgraphs (restricted fractional setting, including $s=1$ as the local case). In this finite-dimensional framework, solutions admit spectral decomposition and exhibit exponential large-time behavior governed by the principal eigenvalue and the spectral gap. In addition, we study positivity improving properties of the associated semigroups for both the lattice and Dirichlet evolutions.

math.AP↗

arXiv2026

Online Decision Making with Fairness over Time

Online platforms increasingly rely on sequential decision-making algorithms to allocate resources, match users, or control exposure, while facing growing pressure to ensure fairness over time. We study a general online decision-making framework in which a platform repeatedly makes decisions from possibly non-convex and discrete feasible sets, such as indivisible assignments or assortment choices, to maximize accumulated reward. Importantly, these decisions must jointly satisfy a set of general, $m$-dimensional, potentially unbounded but convex global constraints, which model diverse long-term fairness goals beyond simple budget caps. We develop a primal-dual algorithm that interprets fairness constraints as dynamic prices and updates them online based on observed outcomes. The algorithm is simple to implement, requiring only the solution of perturbed local optimization problems at each decision step. Under the standard random permutation model, we show that our method achieves $\tilde{O}(\sqrt{mT})$ regret in expected reward while guaranteeing $O(\sqrt{mT})$ violation of long-term fairness constraints deterministically over a horizon of $T$ steps. To capture realistic demand patterns such as periodicity or perturbation, we further extend our guarantees to a grouped random permutation model.

math.OC↗

arXiv2026

Evaluation of real-space second Chern number using the kernel polynomial method

We evaluate the real-space second Chern number of four-dimensional Chern insulators using the kernel polynomial method. Our calculations are performed on a four-dimensional system with $30^4$ sites, and the numerical results agree well with theoretical expectations. Moreover, we show that the method is capable of capturing the disorder effects. This is evidenced by the phase diagram obtained for disordered systems, which agrees well with predictions from the self-consistent Born approximation. Furthermore, we extend the method to six dimensions and perform an exploratory real-space calculation of the third Chern number. Although finite-size effects prevent full quantization, the numerical results show qualitative agreement with theoretical expectations. The study represents a step forward in the real-space characterization of higher-dimensional topological phases.

cond-mat.mes-hall↗

arXiv2026

Correlation between 2D Square Ice and 3D Bulk Ice by Critical Crystallization Pressure

Low-dimensional ice trapped in nanocapillaries is a fascinating phenomenon and is ubiquitous in our daily lives. As a decisive factor of the confinement effect, the size of nanocapillary significantly affects the critical crystallization pressure and crystalline structure, especially for multi-layered ices. By choosing square ice as a typical two-dimensional (2D) multi-layered ice pattern and using all-atom molecular dynamics simulations, we further unveil the variation mechanism of critical crystallization pressure with the nanocapillary size. The results show a strong dependence of the critical crystallization pressure on the size of the graphene sheet for monolayer, bilayer, and trilayer square ice. The quasi-macroscopic crystallization pressure, the actual pressure of water molecules, and the freezable region between them are all strongly dependent on the nanocapillary width. As the size of the capillary becomes larger in all three directions, the critical crystallization pressure converges to the true macroscopic crystallization pressure, which is very close to the value of the crystallization pressure for bulk ice. A direct correlation is established between 2D square ice and three-dimensional (3D) bulk ice by the critical crystallization pressure. There is an unfreezable threshold for crystallizing spontaneously in practice when the quasi-macroscopic crystallization pressure is equal to the actual pressure, which can explain the limit of nanocapillary width for multi-layered ice.

physics.chem-ph↗

arXiv2026

MTDrive: Multi-turn Interactive Reinforcement Learning for Autonomous Driving

Trajectory planning is a core task in autonomous driving, requiring the prediction of safe and comfortable paths across diverse scenarios. Integrating Multi-modal Large Language Models (MLLMs) with Reinforcement Learning (RL) has shown promise in addressing "long-tail" scenarios. However, existing methods are constrained to single-turn reasoning, limiting their ability to handle complex tasks requiring iterative refinement. To overcome this limitation, we present MTDrive, a multi-turn framework that enables MLLMs to iteratively refine trajectories based on environmental feedback. MTDrive introduces Multi-Turn Group Relative Policy Optimization (mtGRPO), which mitigates reward sparsity by computing relative advantages across turns. We further construct an interactive trajectory understanding dataset from closed-loop simulation to support multi-turn training. Experiments on the NAVSIM benchmark demonstrate superior performance compared to existing methods, validating the effectiveness of our multi-turn reasoning paradigm. Additionally, we implement system-level optimizations to reduce data transfer overhead caused by high-resolution images and multi-turn sequences, achieving 2.5x training throughput. Our data, models, and code will be made available soon.

cs.RO↗

arXiv2026

A Two-Stage GPU Kernel Tuner Combining Semantic Refactoring and Search-Based Optimization

GPU code optimization is a key performance bottleneck for HPC workloads as well as large-model training and inference. Although compiler optimizations and hand-written kernels can partially alleviate this issue, achieving near-hardware-limit performance still relies heavily on manual code refactoring and parameter tuning. Recent progress in LLM-agent-based kernel generation and optimization has been reported, yet many approaches primarily focus on direct code rewriting, where parameter choices are often implicit and hard to control, or require human intervention, leading to unstable performance gains. This paper introduces a template-based rewriting layer on top of an agent-driven iterative loop: kernels are semantically refactored into explicitly parameterizable templates, and template parameters are then optimized via search-based autotuning, yielding more stable and higher-quality speedups. Experiments on a set of real-world kernels demonstrate speedups exceeding 3x in the best case. We extract representative CUDA kernels from SGLang as evaluation targets; the proposed agentic tuner iteratively performs templating, testing, analysis, and planning, and leverages profiling feedback to execute constrained parameter search under hardware resource limits. Compared to agent-only direct rewriting, the template-plus-search design significantly reduces the randomness of iterative optimization, making the process more interpretable and enabling a more systematic approach toward high-performance configurations. The proposed method can be further extended to OpenCL, HIP, and other backends to deliver automated performance optimization for real production workloads.

cs.CL↗

arXiv2026

Training-Free In-Context Forensic Chain for Image Manipulation Detection and Localization

Advances in image tampering pose serious security threats, underscoring the need for effective image manipulation localization (IML). While supervised IML achieves strong performance, it depends on costly pixel-level annotations. Existing weakly supervised or training-free alternatives often underperform and lack interpretability. We propose the In-Context Forensic Chain (ICFC), a training-free framework that leverages multi-modal large language models (MLLMs) for interpretable IML tasks. ICFC integrates an objectified rule construction with adaptive filtering to build a reliable knowledge base and a multi-step progressive reasoning pipeline that mirrors expert forensic workflows from coarse proposals to fine-grained forensics results. This design enables systematic exploitation of MLLM reasoning for image-level classification, pixel-level localization, and text-level interpretability. Across multiple benchmarks, ICFC not only surpasses state-of-the-art training-free methods but also achieves competitive or superior performance compared to weakly and fully supervised approaches.

cs.CV↗

arXiv2026

Light-enhanced nonlinear Hall effect

It is well known that a nontrivial Chern number results in quantized Hall conductance. What is less known is that, generically, the Hall response can be dramatically different from its quantized value in materials with broken inversion symmetry. This stems from the leading Hall contribution beyond the linear order, known as the Berry curvature dipole (BCD). While the BCD is in principle always present, it is typically very small outside of a narrow window close to a topological transition and is thus experimentally elusive without careful tuning of external fields, temperature, or impurities. In this work, we transcend this challenge by devising optical driving and quench protocols that enable practical and direct access to large BCD and nonlinear Hall responses. Varying the amplitude of an incident circularly polarized laser drives a topological transition between normal and Chern insulator phases, and importantly allows the precise unlocking of nonlinear Hall currents comparable to or larger than the linear Hall contributions. This strong BCD engineering is even more versatile with our two-parameter quench protocol, as demonstrated in our experimental proposal. Our predictions are expected to hold qualitatively across a broad range of Hall materials, thereby paving the way for the controlled engineering of nonlinear electronic properties in diverse media.

cond-mat.mes-hall↗

arXiv2026

SPARK: Safe Protective and Assistive Robot Kit

This paper introduces the Safe Protective and Assistive Robot Kit (SPARK), a comprehensive benchmark designed to ensure safety in humanoid autonomy and teleoperation. Humanoid robots pose significant safety risks due to their physical capabilities of interacting with complex environments. The physical structures of humanoid robots further add complexity to the design of general safety solutions. To facilitate safe deployment of complex robot systems, SPARK can be used as a toolbox that comes with state-of-the-art safe control algorithms in a modular and composable robot control framework. Users can easily configure safety criteria and sensitivity levels to optimize the balance between safety and performance. To accelerate humanoid safety research and development, SPARK provides simulation benchmarks that compare safety approaches in a variety of environments, tasks, and robot models. Furthermore, SPARK allows quick deployment of synthesized safe controllers on real robots. For hardware deployment, SPARK supports Apple Vision Pro (AVP) or a Motion Capture System as external sensors, while offering interfaces for seamless integration with alternative hardware setups at the same time. This paper demonstrates SPARK's capability with both simulation experiments and case studies with a Unitree G1 humanoid robot. Leveraging these advantages of SPARK, users and researchers can significantly improve the safety of their humanoid systems as well as accelerate relevant research. The open source code is available at: https://github.com/intelligent-control-lab/spark.

cs.RO↗

arXiv2026

Fourier-Jacobi models for real symplectic-metaplectic groups: the basic case

In this paper, we generalize the method of Gan-Ichino and Atobe in [GI16][A18] to the field of real numbers and prove the basic tempered case of the local Gan-Gross-Prasad conjecture for Fourier-Jacobi models of symplectic-metaplectic groups, based on the tempered case of the conjecture for Bessel models proved in [CL22] by Chen-Luo.

math.NT↗

arXiv2026

Light-induced half-quantized Hall effect and axion insulator

Motivated by the recent experimental realization of the half-quantized Hall effect phase in a three-dimensional (3D) semi-magnetic topological insulator [M. Mogi et al., Nature Physics 18, 390 (2022)], we propose a scheme for realizing the half-quantized Hall effect and axion insulator in experimentally mature 3D topological insulator heterostructures. Our approach involves optically pumping and/or magnetically doping the topological insulator surface, such as to break time reversal and gap out the Dirac cones. By toggling between left and right circularly polarized optical pumping, the sign of the half-integer Hall conductance from each of the surface Dirac cones can be controlled, such as to yield half-quantized ($0+1/2$), axion ($-1/2+1/2=0$), and Chern ($1/2+1/2=1$) insulator phases. We substantiate our results based on detailed band structure and Berry curvature numerics on the Floquet Hamiltonian in the high-frequency limit. Our paper showcases how topological phases can be obtained through mature experimental approaches such as magnetic layer doping and circularly polarized laser pumping and opens up potential device applications such as a polarization chirality-controlled topological transistor.

cond-mat.mes-hall↗

arXiv2025

Moiré-modulated $Γ$ valley in twisted bilayer and twisted double-bilayer MoTe$_2$

Twisted MoTe$_2$ hosts intriguing correlated quantum phenomena including the fractional quantum anomalous Hall effect in twisted bilayer (t-BL) MoTe$_2$ near 3.7$^\circ$, which is sensitive to the twist angle and moiré superlattices. Here, we directly visualize the twist-angle-modulated electronic structure of t-BL and twisted double-bilayer (t-DBL) near this critical angle. We find that the moiré superlattice not only modifies the relative energy between $Γ$ and K valleys in t-BL MoTe$_2$, but also strongly reconstructs the $Γ$ valley for both t-BL and t-DBL. Specifically, the deep $p_z$-derived band at $Γ$ exhibits a distinct splitting that systematically varies with increasing twist angle. Theoretical analysis suggests that this modulation arises from the twist-angle-dependent lattice relaxation, especially interfacial corrugations. Our work directly visualizes the moiré-modulated electronic structure and provides key spectroscopic information of lattice relaxation and interlayer interactions underlying the physics of twisted MoTe$_2$.

cond-mat.str-el↗

arXiv2025

Alchemist: Unlocking Efficiency in Text-to-Image Model Training via Meta-Gradient Data Selection

Recent advances in Text-to-Image (T2I) generative models, such as Imagen, Stable Diffusion, and FLUX, have led to remarkable improvements in visual quality. However, their performance is fundamentally limited by the quality of training data. Web-crawled and synthetic image datasets often contain low-quality or redundant samples, which lead to degraded visual fidelity, unstable training, and inefficient computation. Hence, effective data selection is crucial for improving data efficiency. Existing approaches rely on costly manual curation or heuristic scoring based on single-dimensional features in Text-to-Image data filtering. Although meta-learning based method has been explored in LLM, there is no adaptation for image modalities. To this end, we propose **Alchemist**, a meta-gradient-based framework to select a suitable subset from large-scale text-image data pairs. Our approach automatically learns to assess the influence of each sample by iteratively optimizing the model from a data-centric perspective. Alchemist consists of two key stages: data rating and data pruning. We train a lightweight rater to estimate each sample's influence based on gradient information, enhanced with multi-granularity perception. We then use the Shift-Gsampling strategy to select informative subsets for efficient model training. Alchemist is the first automatic, scalable, meta-gradient-based data selection framework for Text-to-Image model training. Experiments on both synthetic and web-crawled datasets demonstrate that Alchemist consistently improves visual quality and downstream performance. Training on an Alchemist-selected 50% of the data can outperform training on the full dataset.

cs.CV↗

arXiv2025

Proper constituent gluon mass as the final piece to construct hybrid mesons

In this letter, we propose that a proper constituent gluon mass $m_g$=450 MeV can be applied to identify the hybrids composed of quarks and gluons. By investigating the spectra and decay widths of the light hybrids $(q\bar{q}g)$ with $J^P=1^{-+}$, we find the $π_1(1600)$ and $η_1(1855)$ may not be explained as $1^{-+}$ hybrids, simultaneously, and the $η_1(1855)$ observed by BESIII may not be a hybrid. In addition, we predict an existence of a hybrid $η_1(1640)$, which can be verified by searching the $a_1(1260)π$ channel. Moreover, we suggest the $K_1(1270)\bar{K}$ and $K_1(1270)π$ as the golden channels to search for an isospin-0 and an isospin-$\frac{1}{2}$ hybrids, respectively.

hep-ph↗