PAPER EVIDENCE RANKING
基于PaperMiner证据链动态评分。缺失证据保持未知，不等同于零分；不同任务论文不宜只凭总分直接比较。
| 排名 | 学校 | 得分 | 论文数 | 覆盖率 |
|---|---|---|---|---|
| 1 | 上海交通大学 DenseStep2M: A Scalable, Training-Free Pipeline for Dense Instructional Video Annotation · Efficient One-Step Diffusion Restoration Model with Compact Token Compression and Linear Attention · ReWorld: Multi-Dimensional Reward Modeling for Embodied World Models | 68.2 | 274 | 92% |
| 2 | Peking University Spiking Variational Graph Representation Inference for Video Summarization · VLA-ReID: Video-Level Association for Re-Identification in Multi-Object Tracking with Highly Similar Objects · HoloTime: Taming Video Diffusion Models for Panoramic 4D Scene Generation | 67.57 | 227 | 93% |
| 3 | University of Science and Technology of China More performant and scalable: Rethinking contrastive vision-language pre-training of radiology in the LLM era · CFSR: Geometry-Conditioned Shadow Removal via Physical Disentanglement · ReWorld: Multi-Dimensional Reward Modeling for Embodied World Models | 67.29 | 183 | 94% |
| 4 | The Chinese University of Hong Kong OARS: Process-Aware Online Alignment for Generative Real-World Image Super-Resolution · Multi-stage image denoising with the wavelet transform · FTDMamba: Frequency-Assisted Temporal Dilation Mamba for Unmanned Aerial Vehicle Video Anomaly Detection | 67.04 | 145 | 92% |
| 5 | Harbin Institute of Technology Spiking Variational Graph Representation Inference for Video Summarization · HoloTime: Taming Video Diffusion Models for Panoramic 4D Scene Generation · UDPNet: Unleashing Depth-based Priors for Robust Image Dehazing | 66.95 | 119 | 90% |
| 6 | Tsinghua University ReWorld: Multi-Dimensional Reward Modeling for Embodied World Models · Fast label-free point-scanning super-resolution imaging for endoscopy · Fi-Gaussian: Frequency-Aware Implicit Gaussian Splatting for Single Image Dehazing | 66.85 | 321 | 92% |
| 7 | Zhejiang University DyST-XL: Dynamic Layout Planning and Content Control for Compositional Text-to-Video Generation · PBE-UNet: A light weight Progressive Boundary-Enhanced U-Net with Scale-Aware Aggregation for Ultrasound Image Segmentation · MLFFM-SegDiff: A Multi-Level Feature Fusion Diffusion Model for Skin Lesion Segmentation | 66.83 | 246 | 94% |
| 8 | Sun Yat-sen University DepthAnything and SAM for UIE: Exploring Large Model Information Contributes to Underwater Image Restoration · CFSR: Geometry-Conditioned Shadow Removal via Physical Disentanglement · SparseGS-W: Sparse-View 3D Gaussian Splatting in the Wild with Generative Priors | 66.68 | 76 | 93% |
| 9 | Nanyang Technological University REprompt: Prompt Generation for Intelligent Software Development Guided by Requirements Engineering · TCPFormer: Learning Temporal Correlation with Implicit Pose Proxy for 3D Human Pose Estimation · Enabling AI-Generated Content (AIGC) Services in Wireless Edge Networks | 66.56 | 165 | 90% |
| 10 | Fudan University MedQ-UNI: Toward Unified Medical Image Quality Assessment and Restoration via Vision-Language Modeling · Can Nano Banana 2 Replace Traditional Image Restoration Models? An Evaluation of Its Performance on Image Restoration Tasks · Dynamic Differential Linear Attention: Enhancing Linear Diffusion Transformer for High-Quality Image Generation | 66.5 | 154 | 91% |
| 11 | National University of Singapore SinAE: A Single-Architecture Flow-Matching Autoencoder for Cross-Domain Atomic Systems · SRDiffusion: Accelerate Video Diffusion Inference via Sketching-Rendering Cooperation · Crisis-Bench: Benchmarking Strategic Ambiguity and Reputation Management in Large Language Models | 66.48 | 147 | 93% |
| 12 | Nanjing University From Zero to Detail: Deconstructing Ultra-High-Definition Image Restoration from Progressive Spectral Perspective · GAP-URGENET: A GENERATIVE-PREDICTIVE FUSION FRAMEWORK FOR UNIVERSAL SPEECH ENHANCEMENT · OLATverse: A Large-scale Real-world Object Dataset with Precise Lighting Control | 66.23 | 89 | 92% |
| 13 | University of California Imagine a City: CityGenAgent for Procedural 3D City Generation · Visual Alignment of Medical Vision-Language Models for Grounded Radiology Report Generation · A Design Study on Voice-based Interaction for Immersive Network Visualization and Analysis | 66.18 | 223 | 94% |
| 14 | The University of Hong Kong Imagine a City: CityGenAgent for Procedural 3D City Generation · ARTEMIS: Autoregressive End-to-End Trajectory Planning with Mixture of Experts for Autonomous Driving · Self-Evaluation Unlocks Any-Step Text-to-Image Generation | 66.18 | 80 | 92% |
| 15 | The Hong Kong University of Science and Technology Dynamic Differential Linear Attention: Enhancing Linear Diffusion Transformer for High-Quality Image Generation · Crisis-Bench: Benchmarking Strategic Ambiguity and Reputation Management in Large Language Models · Exploring the Design Space of Reward Backpropagation for Flow Matching | 66.15 | 162 | 94% |
| 16 | The Hong Kong Polytechnic University Imagine a City: CityGenAgent for Procedural 3D City Generation · Spatial-Frequency Attention for Image Denoising · LLMs for Game Theory: Entropy-Guided In-Context Learning and Adaptive CoT Reasoning | 66.14 | 66 | 92% |
| 17 | University of Chinese Academy of Sciences LibScan: Smart Contract Library Misuse Detection with Iterative Feedback and Static Verification · Spike2Former: Efficient Spiking Transformer for High-performance Image Segmentation · UniSurg: A Video-Native Foundation Model for Universal Understanding of Surgical Videos | 66.13 | 139 | 92% |
| 18 | University of Electronic Science and Technology of China Spiking Variational Graph Representation Inference for Video Summarization · VLA-ReID: Video-Level Association for Re-Identification in Multi-Object Tracking with Highly Similar Objects · From Local Windows to Adaptive Candidates via Individualized Exploratory: Rethinking Attention for Image Super-Resolution | 66.13 | 67 | 94% |
| 19 | Northwestern Polytechnical University DCEvo: Discriminative Cross-Dimensional Evolutionary Learning for Infrared and Visible Image Fusion · Does YOLO Really Need to See Every Training Image in Every Epoch? · Multi-stage image denoising with the wavelet transform | 66.07 | 49 | 92% |
| 20 | Wuhan University REprompt: Prompt Generation for Intelligent Software Development Guided by Requirements Engineering · GeoMAR: Unleashing Geometrically Aligned Features for Masked Autoregressive Blind Face Restoration · Spatial-Spectral Adaptive Fidelity and Noise Prior Reduction Guided Hyperspectral Image Denoising⋆ | 66.06 | 78 | 92% |
| 21 | University of Macau Multi-stage image denoising with the wavelet transform · A Cosine Network for Image Super-Resolution · Spatial-Spectral Adaptive Fidelity and Noise Prior Reduction Guided Hyperspectral Image Denoising⋆ | 66.01 | 37 | 92% |
| 22 | Beihang University LDFE: Laplacian Decoupled Feature Enhancement Block for Dual-Stream CNN-based RGB-IR Object Detection · FS-Diff: Semantic Guidance and Clarity-Aware Simultaneous Multimodal Image Fusion and Super-Resolution · NTIRE 2026 Challenge on Efficient Low Light Image Enhancement: Methods and Results | 65.78 | 84 | 92% |
| 23 | Technical University of Munich UniSurg: A Video-Native Foundation Model for Universal Understanding of Surgical Videos · Unified Video-Action Joint Denoising for Dexterous Action and Data Generation · An Empirical Study of Sampling Hyperparameters in Diffusion-Based Super-Resolution | 65.69 | 86 | 94% |
| 24 | Xidian University IllumFlow: Illumination-Adaptive Low-Light Enhancement via Conditional Rectified Flow and Retinex Decomposition · PointDiffuse: A Dual-Conditional Diffusion Model for Enhanced Point Cloud Semantic Segmentation · ANCHOR: Agentic Noise Creation Framework for Human Simulation and Denoising Recommendation | 65.69 | 50 | 92% |
| 25 | Beijing Institute of Technology MM-DETR: An Efficient Multimodal Detection Transformer with Mamba-Driven Dual-Granularity Fusion and Frequency-Aware Modality Adapters · Compressed-Domain-Aware Online Video Super-Resolution · Video Summarization using Denoising Diffusion Probabilistic Model | 65.59 | 63 | 92% |
| 26 | Dalian University of Technology DCEvo: Discriminative Cross-Dimensional Evolutionary Learning for Infrared and Visible Image Fusion · A super-resolution reconstruction method for lightweight building images based on an expanding feature modulation network · From Zero to Detail: Deconstructing Ultra-High-Definition Image Restoration from Progressive Spectral Perspective | 65.57 | 42 | 92% |
| 27 | Hefei University of Technology AMIF: Authorizable Medical Image Fusion Model with Built-in Authentication · DNA: Dual-stage Native Attribution for Generated Image Source Tracing · Dynamic Spectral Denoising with Global-Context Attention for Multi-Behavior Recommendation | 65.51 | 31 | 94% |
| 28 | Johns Hopkins University Seeing Through the MIRAGE: Evaluating Multimodal Retrieval Augmented Generation · TSCnet: A Text-driven Semantic-level Controllable Framework for Customized Low-Light Image Enhancement · SAW: Toward a Surgical Action World Model via Controllable and Scalable Video Generation | 65.48 | 73 | 93% |
| 29 | Tongji University ARTEMIS: Autoregressive End-to-End Trajectory Planning with Mixture of Experts for Autonomous Driving · INSTANCERSR: REAL-WORLD SUPER-RESOLUTION VIA INSTANCE-AWARE REPRESENTATION ALIGNMENT · Learning from Rendering: Realistic and Controllable Extreme Rainy Image Synthesis for Autonomous Driving Simulation | 65.35 | 61 | 92% |
| 30 | Xi’an Jiaotong University IllumFlow: Illumination-Adaptive Low-Light Enhancement via Conditional Rectified Flow and Retinex Decomposition · SAMIC: A Lightweight Semantic-Aware Mamba for Efficient Perceptual Image Compression · Continuous Expert Assembly: Instance-Conditioned Low-Rank Residuals for All-in-One Image Restoration | 65.33 | 45 | 92% |
| 31 | Massachusetts Institute of Technology Dark matter searches with a 13 meV threshold superconducting sensor array · DiffUS: Differentiable Ultrasound Rendering from Volumetric Imaging · Best Arm Identification with LLM Judges and Limited Human Audits | 65.29 | 138 | 93% |
| 32 | Northeastern University PBE-UNet: A light weight Progressive Boundary-Enhanced U-Net with Scale-Aware Aggregation for Ultrasound Image Segmentation · Generating Multimodal Images with GAN: Integrating Text, Image, and Style · Methodological Frontiers in 21-cm Intensity Mapping: the Treatment of Systematics and Foreground Contamination | 65.29 | 69 | 95% |
| 33 | Carnegie Mellon University LLM-based Multimodal Feedback Produces Equivalent Learning and Better Student Perceptions than Educator Feedback · Hearsay: Vision-Language Medical Diagnoses Without an Image · What Must Generalist Agents Remember? | 65.27 | 105 | 94% |
| 34 | City University of Hong Kong Joint Semantic and Rendering Enhancements in 3D Gaussian Modeling with Anisotropic Local Encoding · From Local Indices to Global Identifiers: Generative Reranking for Recommender Systems via Global Action Space · The geography of novel and atypical research | 65.27 | 70 | 93% |
| 35 | Southeast University ARTEMIS: Autoregressive End-to-End Trajectory Planning with Mixture of Experts for Autonomous Driving · Prior-guided Hierarchical Instance–pixel Contrastive Learning for Ultrasound Speckle Noise Suppression · ScanFocus: A Coarse-to-Fine Framework for Spatio-Temporal Video Grounding | 65.23 | 67 | 93% |
| 36 | Huazhong University of Science and Technology High-resolution Photo Enhancement in Real-time: A Laplacian Pyramid Network · Rectification Reimagined: A Unified Mamba Model for Image Correction and Rectangling with Prompts · DEPTH-SYNERGIZED MAMBA MEETS MEMORY EXPERTS FOR ALL-DAY IMAGE REFLECTION SEPARATION | 65.18 | 65 | 92% |
| 37 | Stanford University ENCORE: Efficient Noise Context-Aware Representation for Low-Dose CT Denoising · Perceptual 3D Simulation With Physical World Modeling · Sliding Window Recurrences for Sequence Models | 65.14 | 142 | 93% |
| 38 | Cornell University MovieRecapsQA: A Multimodal Open-Ended Video Question-Answering Benchmark · Complex Image Generation SwinTransformer Network for Audio Denoising · Propius: A Platform for Collaborative Machine Learning across the Edge and the Cloud | 65.14 | 74 | 96% |
| 39 | University College Dialogue Act Patterns in GenAI-Mediated L2 Oral Practice: A Sequential Analysis of Learner–Chatbot Interactions · AI Feedback Enhances Community-Based Content Moderation through Engagement with Counterarguments · WarpGAN: Warping-Guided 3D GAN Inversion with Style-Based Novel View Inpainting | 65.12 | 80 | 92% |
| 40 | Nankai University Joint Semantic and Rendering Enhancements in 3D Gaussian Modeling with Anisotropic Local Encoding · RESTORE TEXT FIRST, ENHANCE IMAGE LATER: TWO-STAGE SCENE TEXT IMAGE SUPERRESOLUTION WITH GLYPH STRUCTURE GUIDANCE · Devil is in the Uniformity: Exploring Diverse Learners within Transformer for Image Restoration | 65.1 | 40 | 90% |
| 41 | University of Cambridge Fast label-free point-scanning super-resolution imaging for endoscopy · Methodological Frontiers in 21-cm Intensity Mapping: the Treatment of Systematics and Foreground Contamination · DECOMPOSING PRIVATE IMAGE GENERATION VIA COARSE-TO-FINE WAVELET MODELING | 65.07 | 88 | 92% |
| 42 | Beijing University of Posts and Telecommunications Boosting Robustness for All-Weather Self-Supervised Depth Estimation in Autonomous Driving · PIRA: Pan-CDN Intra-video Resource Adaptation for Short Video Streaming · SpaceRipple: Lightweight Semantic Delivery for Mission-Oriented LEO Earth Observation Satellite Networks | 65.07 | 54 | 94% |
| 43 | Indian Institute of Technology Low-Light Image Enhancement Using Gamma Learning And Attention-Enabled Encoder-Decoder Networks · OphEdit: Training-Free Text-Guided Editing of Ophthalmic Surgical Videos · Temporal Object-Aware Vision Transformer for Few-Shot Video Object Detection | 65.06 | 76 | 96% |
| 44 | The University of Tokyo RealX3D: A Physically-Degraded 3D Benchmark for Multi-view Visual Restoration and Reconstruction · QueenVIS: Rethinking Image-Only Training for Video Instance Segmentation via Query Enrichment · Video-Mirai: Autoregressive Video Diffusion Models Need Foresight | 65.06 | 69 | 93% |
| 45 | Shenzhen University M3SR: Multi-Scale Multi-Perceptual Mamba for Efficient Spectral Reconstruction · WeatherRemover: All-in-one Adverse Weather Removal with Multi-scale Feature Map Compression · Continuous Expert Assembly: Instance-Conditioned Low-Rank Residuals for All-in-One Image Restoration | 65.06 | 49 | 92% |
| 46 | New York University VTaMo: Video-Text Alignment Model for Sign Language Translation · EMBOMATRIX: A SCALABLE TRAINING-GROUND FOREMBODIED DECISION-MAKING · ESCA: Enabling Seamless Codec Avatar Execution through Algorithm and Hardware Co-Optimization for Virtual Reality | 65.02 | 85 | 92% |
| 47 | Imperial College Crisis-Bench: Benchmarking Strategic Ambiguity and Reputation Management in Large Language Models · Hybrid Belief–Reinforcement Learning for Efficient Coordinated Spatial Exploration · SELF-SUPERVISED SPATIAL AND ZERO-SHOT ANGULAR SUPER-RESOLUTION BY SPATIAL-ANGULAR IMPLICIT REPRESENTATION FOR ROTATING-VIEW SNR-EFFICIENT DIFFUSION MRI | 65.02 | 85 | 92% |
| 48 | Purdue University ViLaD: A Large Vision Language Diffusion Framework for End-to-End Autonomous Driving · ENABLING DYNAMIC SPARSITY IN QUANTIZED LLM INFERENCE · FathomGPT: A Natural Language Interface for Interactively Exploring Ocean Science Data | 65.02 | 57 | 93% |
| 49 | Nanjing University of Science and Technology STCDiT: Spatio-Temporally Consistent Diffusion Transformer for High-Quality Video Super-Resolution · Decoupling Continual Semantic Segmentation · Devil is in the Uniformity: Exploring Diverse Learners within Transformer for Image Restoration | 65.01 | 44 | 92% |
| 50 | Texas A&M University UNCERTAINTY MATTERS IN DYNAMIC GAUSSIAN SPLATTING FOR MONOCULAR 4D RECONSTRUCTION · MambaLiteUNet: Cross-Gated Adaptive Feature Fusion for Robust Skin Lesion Segmentation · Super-resolution Imaging of Limited-size Objects | 64.95 | 50 | 92% |
排名用于发现值得进一步核查的论文，不替代同行评审。评分会随新增引用、复现、实验验证、新闻评价和落地证据持续更新。
PaperMiner MCP已开放结构化论文反馈。可提交复现成功或失败、代码可运行或失效、Benchmark纠错、链接纠错和部署结果。
MCP接口：submit_paper_feedback、get_paper_community_feedback
权限：有效PaperMiner账号获配的个人MCP密钥可提交；公共代理和共享只读密钥只能查询。每个账号每天最多提交5条新反馈。