2026-09论文中文摘要
面向文档视觉语言模型中细粒度感知的状态条件视觉证据检索
State-Conditioned Visual Evidence Retrieval for Fine-Grained Perception in Document Vision-Language Models
arXiv 2608.28698 · 2026-09-01永久几何占用下的多组管道路由：问题、基准及经典基线
Multi-Group Pipe Routing under Permanent Geometric Occupancy: Problem, Benchmark, and Classical Baselines
arXiv 2608.28697 · 2026-09-01指令蒸馏：将文本指令作为视觉示例
Instruction Distillation: Text Instructions as Visual Examples
arXiv 2608.28696 · 2026-09-01编织视觉叙事：超越原子视觉匹配的智能体图像束合成
Weaving Visual Narratives: Agentic Image Bundle Composition Beyond Atomic Visual Matching
arXiv 2608.28695 · 2026-09-01SNF-Bench：分离长视野固定相机视频生成中的静态漂移与自然流动
SNF-Bench: Separating Static Drift from Natural Flow in Long-Horizon Fixed-Camera Video Generation
arXiv 2608.28694 · 2026-09-01RoboGesture：面向人形机器人交互的实时语义对齐式伴随语音手势生成
RoboGesture: Real-Time Semantic-aligned Co-Speech Gestures Generation for Humanoid Interaction
arXiv 2608.28693 · 2026-09-01培育林分而非单棵树：联合冠层生成再现树冠羞避现象
Growing a Stand, Not a Tree: Joint Canopy Generation Reproduces Crown Shyness
arXiv 2608.28692 · 2026-09-01防御可穿戴视觉语言模型（VLM）抵御私有属性推断
Defending Wearable VLMs Against Private Attribute Inference
arXiv 2608.28691 · 2026-09-0137至42阶主对角线对称Costas阵列的穷尽普查
An Exhaustive Census of Main-Diagonal Symmetric Costas Arrays of Orders 37-42
arXiv 2608.28690 · 2026-09-01托马斯-费米极限下作为受限克尔模拟体的G(2)胶球玻色星
G(2) Glueball Boson Stars as Restricted Kerr Mimickers in the Thomas--Fermi Limit
arXiv 2608.28688 · 2026-09-01FLM：用于生成式图像压缩的频率感知语言模型
FLM: Frequency-Aware Language Models for Generative Image Compression
arXiv 2608.28687 · 2026-09-01诺维科夫-泊松代数的分类及其应用
Classification of Novikov-Poisson Algebras and Their Applications
arXiv 2608.28686 · 2026-09-01数据多样性，而非频率不变性：一项关于抗压缩深度伪造检测的受控自审计研究
Data Diversity, Not Frequency Invariance: A Controlled and Self-Audited Study of Compression-Robust Deepfake Detection
arXiv 2608.28685 · 2026-09-01改进率失真权衡的分布式语义分割
Distributed Semantic Segmentation With Improved Rate-Distortion Trade-Off
arXiv 2608.28684 · 2026-09-01一种将动态刚度矩阵（DSM）与物理信息神经网络（PINN）相结合以求解特征值问题并分析动态响应的框架
A Framework Integrating the Dynamic Stiffness Matrix with Physics-Informed Neural Networks for Solving Eigenvalue Problems and Analysing Dynamic Response
arXiv 2608.28683 · 2026-09-01《Z[T × C₂]中单位群的特征》的更正
Corrigendum to "A Characterization of the Unit Group in Z[T x C_2]"
arXiv 2608.28682 · 2026-09-01CARD：基于反向扩散一致性的域外MRI分割校准方法
CARD: Calibration via Agreement in Reverse Diffusion for Out-of-Domain MRI Segmentation
arXiv 2608.28681 · 2026-09-01利用SAR、多光谱和高光谱影像进行城市弱势住区的多传感器制图：以阿根廷科尔多瓦为例
Multi-Sensor Mapping of Vulnerable Urban Settlements Using SAR, Multispectral, and Hyperspectral Imagery: A Case Study in Córdoba, Argentina
arXiv 2608.28680 · 2026-09-01低普朗特数下均匀双扩散阶梯的通量
Fluxes through a uniform double-diffusive staircase at low Prandtl number
arXiv 2608.28679 · 2026-09-01评估用于可缩放矢量图形生成的约束迭代优化方法（基于现成的视觉语言模型）
Evaluating Constrained Iterative Refinement for Scalable Vector Graphics Generation with Off-the-Shelf VLMs
arXiv 2608.28678 · 2026-09-01面向未知物理环境地面机器人的认知基础型端侧运行时学习
Cognitively-Grounded On-Device Runtime Learning for Ground Robots in Unknown Physical Environments
arXiv 2608.28677 · 2026-09-01标本馆标签数字化自动化流程
Automated pipeline for herbarium label digitization
arXiv 2608.28676 · 2026-09-01用于视频推理的多智能体自增强强化学习
Multi-Agent Self-Improving Reinforcement Learning for Video Reasoning
arXiv 2608.28675 · 2026-09-01多曝光高动态范围（HDR）成像：像素级与特征级重建方法综述
Multi-exposure HDR Imaging: A Review of Pixel-level and Feature-level Reconstruction Methods
arXiv 2608.28674 · 2026-09-01AdaptAV：基于云端专家模型的自动驾驶汽车视觉模型持续适配
AdaptAV: Continuous Adaption of Vision Models for Autonomous Vehicles Using Cloud-based Oracle
arXiv 2608.28673 · 2026-09-01FrameScope：自动驾驶汽车系统中流主动学习的时序数据估值框架
FrameScope: Temporal Data Valuation for Stream Active Learning in Autonomous Vehicle Systems
arXiv 2608.28672 · 2026-09-01使用孪生神经网络衡量艺术作品与AI生成图像的相似性
Measuring Similarity between Artistic and AI Generated Images using Siamese Neural Networks
arXiv 2608.28671 · 2026-09-01基于BaryCache的内存高效无训练扩散Transformer加速方法
Memory-Efficient Training-Free Acceleration of Diffusion Transformers with BaryCache
arXiv 2608.28670 · 2026-09-01MIRAGE-CAD：面向可执行CAD程序的构造介导多模态生成
MIRAGE-CAD: Construction-Mediated Multimodal Generation of Executable CAD Programs
arXiv 2608.28669 · 2026-09-01GreenBench：针对Apple Silicon上开源大语言模型推理的能效与碳足迹的基准测试
GreenBench: Benchmarking Energy Efficiency and Carbon Footprint of Open-Source LLM Inference on Apple Silicon
arXiv 2608.28667 · 2026-09-01通过度量3D融合理解开放词汇无人机感知中的时间语义稳定性
Understanding Temporal Semantic Stability in Open-Vocabulary UAV Perception through Metric 3D Fusion
arXiv 2608.28665 · 2026-09-01触觉基础模型的潜力
The Potential of Haptic Foundation Models
arXiv 2608.28664 · 2026-09-01开放集牛鼻纹识别：泄漏控制基准与评估协议
Open-Set Cattle Muzzle Identification: A Leakage-Controlled Benchmark and Evaluation Protocol
arXiv 2608.28663 · 2026-09-01FRAC-MAS：一种用于骨折诊断的安全可解释多智能体系统
FRAC-MAS: A Safe and Explainable Multi-Agent System for Fracture Diagnosis
arXiv 2608.28662 · 2026-09-01基于Beta说话人活动先验的神经多通道远场说话人 diarization 与源分离
Neural Multichannel Distant Speaker Diarization and Source Separation with Beta Speaker Activity Prior
arXiv 2608.28661 · 2026-09-01科学方程发现的测试时缩放
Test-Time Scaling for Scientific Equation Discovery
arXiv 2608.28660 · 2026-09-01在系统发育网络中寻找树状子结构：整数线性规划方法及其应用
Finding Tree-Like Substructures in Phylogenetic Networks: ILP Approaches and Their Application
arXiv 2608.28659 · 2026-09-01目标驻留使代价和匿名多智能体路径发现问题成为NP难问题
Goal Staying Makes Sum-of-Costs Anonymous Multi-Agent Path Finding NP-Hard
arXiv 2608.28658 · 2026-09-01三角矩阵的定律
Laws for triangular matrices
arXiv 2608.28657 · 2026-09-01RedLight-VLA：用于驾驶策略中交通规则落地与行为强化的模型
RedLight-VLA: Models for traffic-rule grounding and behavioral emphasis in driving policies
arXiv 2608.28656 · 2026-09-01复环面同态的类型
Type of homomorphisms of complex tori
arXiv 2608.28655 · 2026-09-01三次图与系统发育网络的无传递弧定向
Orientations without transitive arcs for cubic graphs and phylogenetic networks
arXiv 2608.28654 · 2026-09-01本原轴代数中根相等性的一个二维反例
A Two-Dimensional Counterexample to Radical Equality in Primitive Axial Algebras
arXiv 2608.28653 · 2026-09-01用于边缘AI推理加速的广义优化引擎（GOE）
A Generalized Optimization Engine (GOE) for Edge AI Inference Acceleration
arXiv 2608.28652 · 2026-09-01有限域上阿贝尔四维簇的Tate猜想
The Tate conjecture for abelian fourfolds over finite fields
arXiv 2608.28651 · 2026-09-01纯量子$(r,ρ)$局部可恢复码的字母相关界
Alphabet-Dependent Bounds for Pure Quantum $(r,ρ)$-Locally Recoverable Codes
arXiv 2608.28650 · 2026-09-01大型语言模型能否识别转化归因中有意义的接触点？
Can Large Language Models Identify Meaningful Touchpoints in Conversion Attribution?
arXiv 2608.28649 · 2026-09-01语言模型如何选择立场：指令层级的内部表征
How Language Models Choose Sides: Internal Representations of Instruction Hierarchy
arXiv 2608.28648 · 2026-09-01面向领域后训练的自专用教师
Self-Specialized Teachers for Domain Post-Training
arXiv 2608.28647 · 2026-09-01BiasMix-Finance：面向LLM投资组合建议的生成后KYC护栏
BiasMix-Finance: Post-Generation KYC Guardrails for LLM Portfolio Advice
arXiv 2608.28646 · 2026-09-01
