Lingdong Kong

Semantic-Aware, Physics-Informed, Geometry-Grounded Weather Synthesis

Towards Unified World Models for Visual Navigation via Memory-Augmented Planning and Foresight

HA-VLN 2.0: An Open Benchmark and Leaderboard for Human-Aware Navigation in Discrete and Continuous Environments

FLUX: Accelerating Cross-Embodiment Generative Navigation Policies via Rectified Flow and Static-to-Dynamic Learning

NavThinker: Action-Conditioned World Models for Coupled Prediction and Planning in Social Navigation

Don't Overthink with Pixels: Efficient Reasoning for Segmentation

Open-o3 Video: Grounded Video Reasoning with Spatio-Temporal Evidence

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

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

EditMGT: Unleashing Potentials of Masked Generative Transformers in Image Editing

AadSFormer: Adaptive Serialized Transformers for Monocular Semantic Scene Completion from Indoor Environments

ReasonMap: Towards Fine-Grained Visual Reasoning from Transit Maps

EventDrive: Event Cameras for Vision-Language Driving Intelligence

AD-R1: Closed-Loop Reinforcement Learning for End-to-End Autonomous Driving with Impartial World Models

Veila: Scaling Diffusion Models for Panoramic LiDAR Point Cloud Generation from a Single Image

RewardMap: Tackling Sparse Rewards in Fine-Grained Visual Reasoning via Multi-Stage Reinforcement Learning

LiDARCrafter: Dynamic 4D World Modeling from LiDAR Sequences

La La LiDAR: Large-Scale Layout Generation from LiDAR Data

See4D: Pose-Free 4D Generation via Auto-Regressive Video Inpainting

Enhanced Spatiotemporal Consistency for Image-to-LiDAR Data Pretraining

Stairway to Success: An Online Floor-Aware Zero-Shot Object-Goal Navigation Framework via LLM-Driven Coarse-to-Fine Exploration

DanceOPD: On-Policy Generative Field Distillation

Inverse Weather Editing: From Entangled Generative Effects to Decomposed Weather Control

Artic4D: Skeleton-Native Motion Diffusion Across Heterogeneous Rigs

VEF: Velocity-Field Distillation for Capability Synthesis in Flow-Matching Models

AI for Auto-Research: Roadmap & User Guide

3D and 4D World Modeling: A Survey

Agentic World Modeling: Foundations, Capabilities, Laws, and Beyond

Watch, Remember, Reason: Human-View Video Understanding with MLLMs

OmniLiDAR: A Unified Diffusion Framework for Multi-Domain 3D LiDAR Generation

STRIDE: Semantically Stratified Historical Retrieval for LiDAR World Modeling

OneVL: One-Step Latent Reasoning with Vision-Language Explanation

Vision-Language-Action Models for Autonomous Driving: Past, Present, and Future

Language-Conditioned World Modeling for Visual Navigation

OneDay: Towards Full-Day Personalized Embodied Agents

Learning to Remove Lens Flare in Event Camera

Talk2Event: Grounded Understanding of Dynamic Scenes from Event Cameras

VideoLucy: Deep Memory Backtracking for Long Video Understanding

3EED: Ground Everything Everywhere in 3D

MERIT: Multilingual Semantic Retrieval with Interleaved Condition Query

SPIRAL: Semantic-Aware Progressive LiDAR Scene Generation and Understanding

FlexEvent: Towards Flexible Event-Frame Object Detection at Varying Operational Frequencies

Perspective-Invariant 3D Object Detection

Are VLMs Ready for Autonomous Driving? An Empirical Study from the Reliability, Data, and Metric Perspectives

Beyond One Shot, Beyond One Perspective: Cross-View and Long-Horizon Distillation for Better LiDAR Representations

MonoMRN: Monocular Semantic Scene Completion via Masked Recurrent Networks

SafeMap: Robust HD Map Construction from Incomplete Observations

EventFly: Event Camera Perception from Ground to the Sky

LiMoE: Mixture of LiDAR Data Representation Learners from Automotive Scenes

GEAL: Generalizable 3D Object Affordance Learning with Cross-Modal Consistency

SeeGround: See and Ground for Zero-Shot Open-Vocabulary 3D Visual Grounding

PointLoRA: Low-Rank Adaptation with Token Selection for Point Cloud Learning

DynamicCity: Large-Scale 4D Occupancy Generation from Dynamic Scenes

Calib3D: Calibrating Model Preferences for Reliable 3D Scene Understanding

LargeAD: Large-Scale Cross-Sensor Data Pretraining for Autonomous Driving

Multi-Modal Data-Efficient 3D Scene Understanding for Autonomous Driving

FRNet: Frustum-Range Networks for Scalable LiDAR-Based Semantic Segmentation

NUC-Net: Non-Uniform Cylindrical Partition Networks for Efficient LiDAR Semantic Segmentation

Visual Foundation Models Boost Cross-Modal Unsupervised Domain Adaptation for 3D Semantic Segmentation

Is Your LiDAR Placement Optimized for 3D Scene Understanding?

Is Your HD Map Constructor Reliable under Sensor Corruptions?

4D Contrastive Superflows are Dense 3D Representation Learners

Learning to Adapt SAM for Segmenting Cross-Domain Point Clouds

OpenESS: Event-Based Scene Understanding with Open Vocabularies

Multi-Space Alignments Towards Universal LiDAR Segmentation

Unified 3D and 4D Panoptic Segmentation via Dynamic Shifting Networks

Benchmarking and Improving Bird's Eye View Perception Robustness in Autonomous Driving

RoboDepth: Robust Out-of-Distribution Depth Estimation under Corruptions

Segment Any Point Cloud Sequences by Distilling Vision Foundation Models

Unsupervised Video Domain Adaptation for Action Recognition: A Disentanglement Perspective

Towards Label-Free Scene Understanding by Vision Foundation Models

Robo3D: Towards Robust and Reliable 3D Perception against Corruptions

Rethinking Range View Representation for LiDAR Segmentation

UniSeg: A Unified Multi-Modal LiDAR Segmentation Network and the OpenPCSeg Codebase

LaserMix for Semi-Supervised LiDAR Semantic Segmentation

CLIP2Scene: Towards Label-Efficient 3D Scene Understanding by CLIP

ConDA: Unsupervised Domain Adaptation for LiDAR Segmentation via Regularized Domain Concatenation

Benchmarking 3D Robustness to Common Corruptions and Sensor Failure