CVPR 2026
自动驾驶 · 具身智能论文归档

RefAV: Towards Planning-Centric Scenario Mining

RiskProp: Collision-Anchored Self-Supervised Risk Propagation For Early Accident Anticipation

SafeDrive: Fine-Grained Safety Reasoning for End-to-End Driving in a Sparse World

CausalVAD: De-confounding End-to-End Autonomous Driving via Causal Intervention

Scaling-Aware Data Selection for End-to-End Autonomous Driving Systems

ActiveAD: Planning-Oriented Active Learning for End-to-End Autonomous Driving

Recover to Predict: Progressive Retrospective Learning for Variable-Length Trajectory Prediction

GuideFlow: Constraint-Guided Flow Matching for Planning in End-to-End Autonomous Driving

Unifying Language-Action Understanding and Generation for Autonomous Driving

Counterfactual VLA: Self-Reflective Vision-Language-Action Model with Adaptive Reasoning

Monocular Open Vocabulary Occupancy Prediction for Indoor Scenes

Customized Fusion: A Closed-Loop Dynamic Network for Adaptive Multi-Task-Aware Infrared-Visible Image Fusion

SeeGroup: Multi-Layer Depth Estimation of Transparent Surfaces via Self-Determined Grouping

ComPose: A Unified Completion-Pose Framework for Robust Category-Level Object Pose Estimation

CoSMo3D: Open-World Promptable 3D Semantic Segmentation through LLM-Guided Canonical Spatial Modeling

RobotSeg: A Model and Dataset for Segmenting Robots in Image and Video

S^2AM3D: Scale-controllable Part Segmentation of 3D Point Clouds

VGGT-Segmentor: Geometry-Enhanced Cross-View Segmentation

Learning Latent Concepts for Detecting Out-of-Distribution Objects

Optical Flow Matching: Reframing Optical Flow as Continuous Transport Dynamics

SEATrack: Simple, Efficient, and Adaptive Multimodal Tracker

U^2Flow: Uncertainty-Aware Unsupervised Optical Flow Estimation

SDTrack: A Baseline for Event-based Tracking via Spiking Neural Networks

Differentiable Laplacian Matrix Guided Superpixel Segmentation

Dual-level Adapter Boosting Prompt-free Curvilinear Structure Segmentation

CD-Buffer: Complementary Dual-Buffer Framework for Test-Time Adaptation in Adverse Weather Object Detection

DA-Mamba: Learning Domain-Aware State Space Model for Global-Local Alignment in Domain Adaptive Object Detection

Event-based Motion Deblurring with Unpaired Data

Depth Hypothesis Guided Iterative Refinement for Event–Image Monocular Depth Estimation

Event-based Visual Deformation Measurement

From Contrast to Consistency: Rethinking Event-based Continuous-Time Optical Flow Estimation

High-Quality and Efficient Turbulence Mitigation with Events

x^2-Fusion: Cross-Modality and Cross-Dimension Flow Estimation in Event Edge Space

CIGPose: Causal Intervention Graph Neural Network for Whole-Body Pose Estimation

Curvature-Aware Captioning: Leveraging Geodesic Attention for 3D Scene Understanding

SAQN: Semantic-based Adaptive Query Network for 3D Referring Expression Segmentation

V^2-SAM: Marrying SAM2 with Multi-Prompt Experts for Cross-View Object Correspondence

LitePT: Lighter Yet Stronger Point Transformer

Explaining Object Detectors via Collective Contribution of Pixels

Mamba Learns in Context: Structure-Aware Domain Generalization for Multi-Task Point Cloud Understanding

Denoise and Align: Towards Source-Free UDA for Robust Panoramic Semantic Segmentation

TeFlow: Enabling Multi-frame Supervision for Self-Supervised Feed-forward Scene Flow Estimation

Changes in Real Time: Online Scene Change Detection with Multi-View Fusion

Mitigating Objectness Bias and Region-to-Text Misalignment for Open-Vocabulary Panoptic Segmentation

RAVEN: Radar Adaptive Vision Encoders for Efficient Chirp-wise Object Detection and Segmentation

Heuristic Self-Paced Learning for Domain Adaptive Semantic Segmentation under Adverse Conditions

Text-guided Feature Disentanglement for Cross-modal Gait Recognition

3D-Object Perception Transformer (3PT)

RARE: Learn to RAnk and REtrieve for Monocular 3D Object Detection

SABER: Spatially Consistent 3D Universal Adversarial Objects for BEV Detectors

Scene Reconstruction as Mapping Priors for 3D Detection

Unleashing the Power of Chain-of-Prediction for Monocular 3D Object Detection

UniFusion: A Unified Image Fusion Framework with Robust Representation and Source-Aware Preservation

240FPS Stereo Vision from Monocular Mixed Spikes

Missing No More: Dictionary-Guided Cross-Modal Image Fusion under Missing Infrared

SpiderCam: Low-Power Snapshot Depth from Differential Defocus

LiteSense: Lifting Lightweight ToF with RGB for High-Resolution Metric Depth Estimation

The Midas Touch for Metric Depth

PET-DINO: Unifying Visual Cues into Grounding DINO with Prompt-Enriched Training

Better than Average: Spatially-Aware Aggregation of Segmentation Uncertainty Improves Downstream Performance

D-Convexity: A Unified Differentiable Convex Shape Prior via Quasi-Concavity for Data-driven Image Segmentation

Open-Vocabulary Domain Generalization in Urban-Scene Segmentation

Retrieve and Segment: Are a Few Examples Enough to Bridge the Supervision Gap in Open-Vocabulary Segmentation?

Boosting Vision-Language Models Towards Cross-Domain Incremental Object Detection

RaUF: Learning the Spatial Uncertainty Field of Radar

PromptMoE: A Segmentation Refinement Framework Leveraging Mixture of Experts for Improved Prompting

Thermally Activated Dual-Modal Adversarial Clothing against AI Surveillance Systems

Real-World Point Tracking with Verifier-Guided Pseudo-Labeling

Fractal Camouflage: A Bio-Inspired Approach for Multi-Scale Adversarial Attacks in the Infrared Domain

Mocap-2-to-3: Multi-view Lifting for Monocular Motion Recovery with 2D Pretraining

Learning to Track Instance from Single Nature Language Description

Beyond Static Frames: Temporal Aggregate-and-Restore Vision Transformer for Human Pose Estimation

Adaptive Depth Lightweight RGB-T Tracking with Holistic Token Routing

Wavelet-Driven 3D Anomaly Detection under Pose-Agnostic and Sparse-View

GSV2X: Geometry-Aware Uncertainty Modeling and Orthogonal Fusion for Robust Roadside Perception

A Semantically Disentangled Unified Model for Multi-category 3D Anomaly Detection

Beyond Strict Pairing: Arbitrarily Paired Training for High-Performance Infrared and Visible Image Fusion

FreqSIC: Frequency-aware Stereo Image Compression with Bi-directional Checkerboard Context Model

SparseWorld-TC: Trajectory-Conditioned Sparse Occupancy World Model

ProPhy: Progressive Physical Alignment for Dynamic World Simulation

WorldLens: Full-Spectrum Evaluations of Driving World Models in Real World

Inference-time Physics Alignment of Video Generative Models with Latent World Models

4D-RGPT: Toward Region-level 4D Understanding via Perceptual Distillation

Physical Object Understanding with a Physically Controllable World Model

U4D: Uncertainty-Aware 4D World Modeling from LiDAR Sequences

NeoVerse: Enhancing 4D World Model with in-the-wild Monocular Videos

StaMo: Unsupervised Learning of Generalizable Robot Motion from Compact State Representation

PointWorld: Scaling 3D World Models for In-The-Wild Robotic Manipulation

A Frame is Worth One Token: Efficient Generative World Modeling with Delta Tokens

Interpretable Motion-Attentive Maps: Spatio-Temporally Localizing Concepts in Video Diffusion Transformers

NitroGen: An Open Foundation Model for Generalist Gaming Agents

AT-VLA: Adaptive Tactile Injection for Enhanced Feedback Reaction in Vision-Language-Action Models

Structural Action Transformer for 3D Dexterous Manipulation

CLaD: Planning with Grounded Foresight via Cross-Modal Latent Dynamics

SaPaVe: Towards Active Perception and Manipulation in Vision-Language Action Models for Robotics

ForeAct: Steering Your VLA with Efficient Visual Foresight Planning

When Robots Obey the Patch: Universal Transferable Patch Attacks on Vision-Language-Action Models

Towards Human-Like Robot Handwriting via Contour-Aware Generation

BiPreManip: Learning Affordance-Based Bimanual Pre-Manipulation through Anticipatory Collaboration

AVA-VLA: Improving Vision-Language-Action models with Active Visual Attention

Cross-Hand Latent Representation for Vision-Language-Action Models

GeoPredict: Leveraging Predictive Kinematics and 3D Gaussian Geometry for Precise VLA Manipulation

Boosting Vision-Language-Action Finetuning with Feasible Action Neighborhood Prior

CycleManip: Enabling Cycle-based Manipulation via Effective History Perception and Understanding

Efficient Hybrid SE(3)-Equivariant Visuomotor Flow Policy via Spherical Harmonics for Robot Manipulation

GraspALL: Adaptive Structural Compensation from Illumination Variation for Robotic Garment Grasping in Any Low-Light Conditions

EnergyAction: Unimanual to Bimanual Composition with Energy-Based Models

Do You Have Freestyle? Expressive Humanoid Locomotion via Audio Control

InterPrior: Scaling Generative Control for Physics-Based Human-Object Interactions

Unified Number-Free Text-to-Motion Generation Via Flow Matching

CoordSpeaker: Exploiting Gesture Captioning for Coordinated Caption-Empowered Co-Speech Gesture Generation

Geometric Neural Distance Fields for Learning Human Motion Priors

End-to-End Language-Action Model for Humanoid Whole Body Control

MaskAdapt: Learning Flexible Motion Adaptation via Mask-Invariant Prior for Physics-Based Characters

FloodDiffusion: Tailored Diffusion Forcing for Streaming Motion Generation

Action Motifs: Self-Supervised Hierarchical Representation of Human Body Movements

Gallant: Voxel Grid-based Humanoid Locomotion and Local-navigation across 3-D Constrained Terrains

MAGICIAN: Efficient Long-Term Planning with Imagined Gaussians for Active Mapping

Memory-Augmented Scene Understanding and Exploration for Open-World Aerial Object-Goal Navigation

SocialNav: Training Human-Inspired Foundation Model for Socially-Aware Embodied Navigation

Training One Model to Master Cross-Level Agentic Actions via Reinforcement Learning

Multi-Scale Gaussian-Language Map for Zero-shot Embodied Navigation and Reasoning

GeniNav: Generative Model Driven Image-Goal Navigation via Imagination-Guided Consistency Flow Matching

EgoAVU: Egocentric Audio-Visual Understanding

ViterbiPlanNet: Injecting Procedural Knowledge via Differentiable Viterbi for Planning in Instructional Videos

HTNav: A Hybrid Navigation Framework with Tiered Structure for Urban Aerial Vision-and-Language Navigation

OVI-MAP: Open-Vocabulary Instance-Semantic Mapping

STRNet: Visual Navigation with Spatio-Temporal Representation through Dynamic Graph Aggregation

Revisiting Geometric Obfuscation with Dual Convergent Lines for Privacy-Preserving Image Queries in Visual Localization

PoseGAM: Robust Unseen Object Pose Estimation via Geometry-Aware Multi-View Reasoning

FUSER: Feed-Forward Multiview 3D Registration Transformer and SE(3)^N Diffusion Refinement

From Pairs to Sequences: Track-Aware Policy Gradients for Keypoint Detection

Linear Fundamental Matrix Estimation from 7 or 5 Points

Global-Aware Edge Prioritization for Pose Graph Initialization

QuadSync: Quadrifocal Tensor Synchronization via Tucker Decomposition

TESO: Online Tracking of Essential Matrix by Stochastic Optimization

MV-RoMa: From Pairwise Matching into Multi-View Track Reconstruction

Learning Scene Coordinate Reconstruction from Unposed Images via Pose Graph Optimization

Parallel Rigidity Matters for Bundle Adjustment

Simple but Effective Triplet-Based Compression Strategies for Compact Visual Localization

Global Structure-from-Motion Meets Feedforward Reconstruction

From Corners to Fiducial Tags: Revisiting Checkerboard Calibration for Event Cameras

SuP: Sub-cloud Driven Point Cloud Registration

LEADER: Learning Reliable Local-to-Global Correspondences for LiDAR Relocalization

GM-R^2: Generative Matching Learning for Unsupervised Geometric Representation and Registration

Generalized-CVO: Fast and Correspondence-Free Local Point Cloud Registration with Second Order Riemannian Optimization

StreamVLO: Streaming Visual–LiDAR Odometry with Cumulative Drift Compensation

Globally Optimal Pose from Orthographic Silhouettes

Cov2Pose: Leveraging Spatial Covariance for Direct Manifold-aware 6-DoF Object Pose Estimation

Flow4DGS-SLAM: Optical Flow-Guided 4D Gaussian Splatting SLAM

ULF-Loc: Unbiased Landmark Feature for Robust Visual Localization with 3D Gaussian Splatting

Homaloidal parametrization for detecting critical two-view configurations

CoLoR: The Devil is in Scene Coordinate Regression for Large-Scale Visual Localization

AsymLoc: Towards Asymmetric Feature Matching for Efficient Visual Localization

Affine Perspective-Three-Point Problem

VIRD: View-Invariant Representation through Dual-Axis Transformation for Cross-View Pose Estimation

SinGeo: Unlock Single Model's Potential for Robust Cross-View Geo-Localization

GeoAgent: Learning to Geolocate Everywhere with Reinforced Geographic Characteristics

VLM-Loc: Localization in Point Cloud Maps via Vision-Language Models

HOLO: Homography-Guided Pose Estimator Network for Fine-Grained Visual Localization on SD Maps

PiLoT: Neural Pixel-to-3D Registration for UAV-based Ego and Target Geo-localization

Solving Minimal Problems Without Matrix Inversion Using FFT-Based Interpolation

Dark3R: Learning Structure from Motion in the Dark

Visual-RRT: Finding Paths toward Visual-Goals via Differentiable Rendering

Scalable Feature Matching via State Space Modeling and Sparse Correlation

HamiPose: Hamiltonian Optimization for Unsupervised Domain Adaptive Pose Estimation

3DReflecNet: A Large-Scale Dataset for 3D Reconstruction of Reflective, Transparent, and Low-Texture Objects

SPARK: Sim-ready Part-level Articulated Reconstruction with VLM Knowledge

PAI-Bench: A Comprehensive Benchmark For Physical AI

InfiniBench: Infinite Benchmarking for Visual Spatial Reasoning with Customizable Scene Complexity

OccuFly: A 3D Vision Benchmark for Semantic Scene Completion from the Aerial Perspective

VS-Bench: Evaluating VLMs for Strategic Abilities in Multi-Agent Environments

OpenDance: Multimodal Controllable 3D Dance Generation with Large-scale Internet Data

SimScale: Learning to Drive via Real-World Simulation at Scale

Texvent: Asynchronous Event Data Simulation via Text Prompt

The SA-FARI Dataset: Segment Anything in Footage of Animals for Recognition and Identification

RealAppiance: Let High-fidelity Appliance Assets Controllable and Workable as Aligned Real Manauls

Wanderland: Geometrically Grounded Simulation for Open-World Embodied AI

When Robots Should Say ''I Don’t Know'': Benchmarking Abstention in Embodied Question Answering

Towards Generalized Multimodal Homography Estimation

HG-Lane: High-Fidelity Generation of Lane Scenes under Adverse Weather and Lighting Conditions without Re-annotation

Hoi! - A Multimodal Dataset for Force-Grounded, Cross-View Articulated Manipulation

Scalable Trajectory Generation for Whole-Body Mobile Manipulation

SpatialScore: Towards Comprehensive Evaluation for Spatial Intelligence

SpatialTree: How Spatial Intelligence Branches Out in MLLMs

PanoEnv: Exploring 3D Spatial Intelligence in Panoramic Environments with Reinforcement Learning

Task-Driven Implicit Representations for Automated Design of LiDAR Systems

DENALI: A Dataset Enabling Non-Line-of-Sight Spatial Reasoning with Low-Cost LiDARs

HanDyVQA: A Video QA Benchmark for Fine-Grained Hand-Object Interaction Dynamics

RLFTSim: Realistic and Controllable Multi-Agent Traffic Simulation via Reinforcement Learning Fine-Tuning

CARD: A Multi-Modal Automotive Dataset for Dense 3D Reconstruction in Challenging Road Topography

PerpetualWonder: Long-horizon Action-conditioned 4D Scene Generation

Portable Active Learning for Object Detection

Online Data Curation for Object Detection via Marginal Contributions to Dataset-level Average Precision

EgoXtreme: A Dataset for Robust Object Pose Estimation in Egocentric Views under Extreme Conditions

SceneScribe-1M: A Large-Scale Video Dataset with Comprehensive Geometric and Semantic Annotations

What Makes Good Synthetic Training Data for Zero-Shot Stereo Matching?

M3DLayout: A Multi-Source Dataset of 3D Indoor Layouts and Structured Descriptions for 3D Generation

SimRecon: SimReady Compositional Scene Reconstruction from Real Videos

Diffusion-Based sRGB Real Noise Generation via Prompt-Driven Noise Representation Learning

OMG-Bench: A New Challenging Benchmark for Skeleton-based Online Micro Hand Gesture Recognition

MAMMA: Markerless Accurate Multi-person Motion Acquisition

Energy-GS: Image Energy-guided Pose Alignment Gaussian Splatting with redesigned pose gradient flow

4D Primitive-Mâché: Glueing Primitives for Persistent 4D Scene Reconstruction

Efficiently Reconstructing Dynamic Scenes One D4RT at a Time

Natural Human Motion Recovery by Aligning High-Order Temporal Dynamics from Monocular Videos

Residual Primitive Fitting of 3D Shapes with SuperFrusta

RetimeGS: Continuous-Time Reconstruction of 4D Gaussian Splatting

SAM 3D Body: Robust Full-Body Human Mesh Recovery

SmokeSVD: Smoke Reconstruction from A Single View via Progressive Novel View Synthesis and Refinement with Diffusion Models

PPISP: Physically-Plausible Compensation and Control of Photometric Variations in Radiance Field Reconstruction

SAM 3D: 3Dfy Anything in Images

Selfi: Self-improving Reconstruction Engine via 3D Geometric Feature Alignment

Faithful Contouring: Near-Lossless 3D Voxel Representation Free from Iso-surface

Native and Compact Structured Latents for 3D Generation

Chorus: Multi-Teacher Pretraining for Holistic 3D Gaussian Scene Encoding

GaussianFluent: Gaussian Simulation for Dynamic Scenes with Mixed Materials

VGGT-Ω

Evidential Neural Radiance Fields

FILTR: Extracting Topological Features from Pretrained 3D Models

Learning Convex Decomposition via Feature Fields

UnReflectAnything: RGB-Only Highlight Removal by Rendering Synthetic Specular Supervision

Learning Eigenstructures of Unstructured Data Manifolds

MoRe: Motion-aware Feed-forward 4D Reconstruction Transformer

Scal3R: Scalable Test-Time Training for Large-Scale 3D Reconstruction

OmniVGGT: Omni-Modality Driven Visual Geometry Grounded Transformer

E-RayZer: Self-supervised 3D Reconstruction as Spatial Visual Pre-training

Point4Cast: Streaming Dynamic Scene Reconstruction and Forecasting

AMB3R: Accurate Feed-forward Metric-scale 3D Reconstruction with Backend

STAC: Plug-and-Play Spatio-Temporal Aware Cache Compression for Streaming 3D Reconstruction

No Calibration, No Depth, No Problem: Cross-Sensor View Synthesis with 3D Consistency

SLARM: Streaming and Language-Aligned Reconstruction Model for Dynamic Scenes

RnG: A Unified Transformer for Complete 3D Modeling from Partial Observations

tttLRM: Test-Time Training for Long Context and Autoregressive 3D Reconstruction

Geometric-Photometric Event-based 3D Gaussian Ray Tracing

ChronoGS: Disentangling Invariants and Changes in Multi-Period Scenes

TokenGS: Decoupling 3D Gaussian Prediction from Pixels with Learnable Tokens

Hermite Radial Basis Function for Surface Reconstruction via Differentiable Rendering

RF4D:Neural Radar Fields for Novel View Synthesis in Outdoor Dynamic Scenes

ParticleGS: Learning Neural Gaussian Particle Dynamics from Videos for Prior-free Physical Motion Extrapolation

WildRayZer: Self-supervised Large View Synthesis in Dynamic Environments

PhysGM: Large Physical Gaussian Model for Feed-Forward 4D Synthesis

Seeing through Light and Darkness: Sensor-Physics Grounded Deblurring HDR NeRF from Single-Exposure Images and Events

Text-Driven 3D Hand Motion Generation from Sign Language Data

Clay-to-Stone: Phase-wise 3D Gaussian Splatting for Monocular Articulated Hand-Object Manipulation Modeling

TouchDream: 3D Object Completion through Imagined Touch

ArtHOI: Taming Foundation Models for Monocular 4D Reconstruction of Hand-Articulated-Object Interactions

Glove2Hand: Synthesizing Natural Hand-Object Interaction from Multi-Modal Sensing Gloves

IntrinsicWeather: Controllable Weather Editing in Intrinsic Space

DMAligner: Enhancing Image Alignment via Diffusion Model Based View Synthesis

InstantViR: Real-Time Video Inverse Problem Solver with Distilled Diffusion Prior

Unified Vector Floorplan Generation via Markup Representation

Vista4D: Video Reshooting with 4D Point Clouds

MotionCrafter: Dense Geometry and Motion Reconstruction with a 4D VAE

MatLat: Material Latent Space for PBR Texture Generation

EcoSplat: Efficiency-controllable Feed-forward 3D Gaussian Splatting from Multi-view Images

Urban-GS: A Unified 3D Gaussian Splatting Framework for Compact and High-Fidelity Aerial-to-Street Reconstruction

Uni3R: Unified 3D Reconstruction and Semantic Understanding via Generalizable Gaussian Splatting from Unposed Multi-View Images

TokenSplat: Token-aligned 3D Gaussian Splatting for Feed-forward Pose-free Reconstruction

AeroGS: Scale-Aware Gaussian Splatting for Pose-Free Dynamic UAV Scene Reconstruction

TagSplat: Topology-Aware Gaussian Splatting for Dynamic Mesh Modeling and Tracking

BEA-GS: BEyond RAdiance Supervision in 3DGS for Precise Object Extraction

E2EGS: Event-to-Edge Gaussian Splatting for Pose-Free 3D Reconstruction

DirectFisheye-GS: Enabling Native Fisheye Input in Gaussian Splatting with Cross-View Joint Optimization

CUPID: Generative 3D Reconstruction via Joint Object and Pose Modeling

RawMetaDiff: Unlocking Extreme Darkness from Dual-Exposure RAW with Meta-Guided Diffusion

Ov3R: Open-Vocabulary Semantic 3D Reconstruction from RGB Videos

UniPart: Part-Level 3D Generation with Unified 3D Geom–Seg Latents

UniPixie: Unified and Probabilistic 3D Physics Learning via Flow Matching

2D-LFM: Lifting Foundation Model without 3D Supervision

NG-GS: NeRF-guided 3D Gaussian Splatting Segmentation

RINO: Rotation-Invariant Non-Rigid Correspondences

Tracking by Predicting 3-D Gaussians Over Time

ResiHMR: Residual-Limb Aware Single-Image 3D Human Mesh Recovery for Individuals with Limb Loss

EgoPoseFormer v2: Accurate Egocentric Human Motion Estimation for AR/VR

PoseD-Flow: Versatile and Guided Flow Matching Model of Human Pose

PAD-Hand: Physics-Aware Diffusion for Hand Motion Recovery

Human Interaction-Aware 3D Reconstruction from a Single Image

Recovering Physically Plausible Human-Object Interactions from Monocular Videos

TeHOR: Text-Guided 3D Human and Object Reconstruction with Textures

BiMotion: B-spline Motion for Text-guided Dynamic 3D Character Generation

CoRoGS: Contextual Gaussian Splatting for Robust Large-Deviation View Synthesis

Content-Aware Frequency Encoding for Implicit Neural Representations with Fourier-Chebyshev Features

FACE: A Face-based Autoregressive Representation for High-Fidelity and Efficient Mesh Generation

MeshRipple: Structured Autoregressive Generation of Artist-Meshes

NERFIFY: A Multi-Agent Framework for Turning NeRF Papers into Code

One-Shot Flow, Any-Time Frame: A Bidirectional Warping Framework for Event-Based Video Frame Interpolation

PhysSkin: Real-Time and Generalizable Physics-Based Animation via Self-Supervised Neural Skinning

Taming Video Models for 3D and 4D Generation via Zero-Shot Camera Control

gQIR: Generative Quanta Image Reconstruction

Semantic Foam: Unifying Spatial and Semantic Scene Decomposition

Color-Encoded Illumination for High-Speed Volumetric Scene Reconstruction

FISHuman: Fine-grained Single-image 3D Human Reconstruction via Multi-view 4D Remeshing

Does YOLO Really Need to See Every Training Image in Every Epoch?

NuWa: Deriving Lightweight Class-Specific Vision Transformers for Edge Devices

SenCache: Accelerating Diffusion Model Inference via Sensitivity-Aware Caching

Streaming Diffusion Model for Fast Infrared and Visible Video Fusion

INSID3: Training-Free In-Context Segmentation with DINOv3

Featurising Pixels from Dynamic 3D Scenes with Linear In-Context Learners

MARCO: Navigating the Unseen Space of Semantic Correspondence

PR-MaGIC: Prompt Refinement Via Mask Decoder Gradient Flow For In-Context Segmentation

AToken: A Unified Tokenizer for Vision

ANTS: Adaptive Negative Textual Space Shaping for OOD Detection via Test-Time MLLM Understanding and Reasoning

SeaCache: Spectral-Evolution-Aware Cache for Accelerating Diffusion Models

ViT^3: Unlocking Test-Time Training in Vision

Mapping Networks

AVGGT: Rethinking Global Attention for Accelerating VGGT

A Mixed Diet Makes DINO An Omnivorous Vision Encoder

Scaling Dense Event-Stream Pretraining from Visual Foundation Models

Towards Multimodal Domain Generalization with Few Labels

SAMTok: Representing Any Mask with Two Words

Finding Distributed Object-Centric Properties in Self-Supervised Transformers

SigLino: Efficient Multi-Teacher Distillation for Agglomerative Vision Foundation Models

The Invisible Gorilla Effect in Out-of-distribution Detection

CoIn: Coverage and Informativeness-Guided Token Reduction for Efficient Large Multimodal Models

One Layer’s Trash is Another Layer’s Treasure: Adaptive Layer-wise Visual Token Selection in LVLMs

SSM-Aware Token-Efficient VMamba via Adaptive Patch Pruning and Merging for Person Re-Identification

Pluggable Pruning with Contiguous Layer Distillation for Diffusion Transformers

VMonarch: Efficient Video Diffusion Transformers with Structured Attention

Draft and Refine with Visual Experts

ApET: Approximation-Error Guided Token Compression for Efficient VLMs

All Roads Lead to Rome: Incentivizing Divergent Thinking in Vision-Language Models

NAF: Zero-Shot Feature Upsampling via Neighborhood Attention Filtering

ReMoT: Reinforcement Learning with Motion Contrast Triplets

Gated KalmaNet: A Fading Memory Layer through Test-time Ridge Regression

Evidential Transformation Network: Turning Pretrained Models into Evidential Models for Post-hoc Uncertainty Estimation

Dual-branch Distilled Transformer for Efficient Asymmetric UAV Tracking

Verifying Neural Network Robustness with Dual Perturbations

An Efficient Token Compression Framework for Visual Object Tracking

DDiT: Dynamic Patch Scheduling for Efficient Diffusion Transformers

Keep It Frozen: Domain-Routed Conditional Residual Modulation for Multi-Domain Vision Transformers

MSPT: Efficient Large-Scale Physical Modeling via Parallelized Multi-Scale Attention

Mechanisms of Object Localization in Vision–Language Models

THE MORE, THE MERRIER: CONTRASTIVE FUSION FOR HIGHER-ORDER MULTIMODAL ALIGNMENT

Trainable Log-linear Sparse Attention for Efficient Diffusion Transformers

TriLite: Efficient Weakly Supervised Object Localization with Universal Visual Features and Tri-Region Disentanglement

Visual Diffusion Models are Geometric Solvers

What Do Visual Tokens Really Encode? Uncovering Sparsity and Redundancy in Multimodal Large Language Models

When Do Models Actually Decide? Mapping the Layer-Wise Decision Timeline in Pretrained Neural Networks

Inside-Out: Measuring Generalization in Vision Transformers Through Inner Workings

Attend Before Attention: Efficient and Scalable Video Understanding via Autoregressive Gazing

Concept-Guided Fine-Tuning: Steering ViTs away from Spurious Correlations to Improve Robustness