Tai Wang

papers

1,887

total citations

papers (17)

LLaVA-3D: A Simple yet Effective Pathway to Empowering LMMs with 3D Capabilities

ICCV 2025

127

citations

Unified Human-Scene Interaction via Prompted Chain-of-Contacts

ICLR 2024arXiv

101

citations

Monocular 3D Object Detection with Depth from Motion

ECCV 2022arXiv

citations

MV-JAR: Masked Voxel Jigsaw and Reconstruction for LiDAR-Based Self-Supervised Pre-Training

CVPR 2023arXiv

citations

GenNBV: Generalizable Next-Best-View Policy for Active 3D Reconstruction

CVPR 2024arXiv

citations

GeoMIM: Towards Better 3D Knowledge Transfer via Masked Image Modeling for Multi-view 3D Understanding

ICCV 2023arXiv

citations

RoboGround: Robotic Manipulation with Grounded Vision-Language Priors

CVPR 2025arXiv

citations

GENMANIP: LLM-driven Simulation for Generalizable Instruction-Following Manipulation

CVPR 2025arXiv

citations

Rethinking the Embodied Gap in Vision-and-Language Navigation: A Holistic Study of Physical and Visual Disparities

ICCV 2025arXiv

citations

OST-Bench: Evaluating the Capabilities of MLLMs in Online Spatio-temporal Scene Understanding

NEURIPS 2025arXiv

citations

GLEAM: Learning Generalizable Exploration Policy for Active Mapping in Complex 3D Indoor Scene

ICCV 2025arXiv

citations

VFlowOpt: A Token Pruning Framework for LMMs with Visual Information Flow-Guided Optimization

ICCV 2025arXiv

citations

Tai Wang

papers (17)

Cylindrical and Asymmetrical 3D Convolution Networks for LiDAR Segmentation

PointLLM: Empowering Large Language Models to Understand Point Clouds

Scene as Occupancy

MonoDETR: Depth-guided Transformer for Monocular 3D Object Detection

EmbodiedScan: A Holistic Multi-Modal 3D Perception Suite Towards Embodied AI

LLaVA-3D: A Simple yet Effective Pathway to Empowering LMMs with 3D Capabilities

Unified Human-Scene Interaction via Prompted Chain-of-Contacts

Monocular 3D Object Detection with Depth from Motion

MV-JAR: Masked Voxel Jigsaw and Reconstruction for LiDAR-Based Self-Supervised Pre-Training

GenNBV: Generalizable Next-Best-View Policy for Active 3D Reconstruction

GeoMIM: Towards Better 3D Knowledge Transfer via Masked Image Modeling for Multi-view 3D Understanding

RoboGround: Robotic Manipulation with Grounded Vision-Language Priors

GENMANIP: LLM-driven Simulation for Generalizable Instruction-Following Manipulation

Rethinking the Embodied Gap in Vision-and-Language Navigation: A Holistic Study of Physical and Visual Disparities

OST-Bench: Evaluating the Capabilities of MLLMs in Online Spatio-temporal Scene Understanding

GLEAM: Learning Generalizable Exploration Policy for Active Mapping in Complex 3D Indoor Scene

VFlowOpt: A Token Pruning Framework for LMMs with Visual Information Flow-Guided Optimization

papers (17)

Cylindrical and Asymmetrical 3D Convolution Networks for LiDAR Segmentation

PointLLM: Empowering Large Language Models to Understand Point Clouds

Scene as Occupancy

MonoDETR: Depth-guided Transformer for Monocular 3D Object Detection

EmbodiedScan: A Holistic Multi-Modal 3D Perception Suite Towards Embodied AI

LLaVA-3D: A Simple yet Effective Pathway to Empowering LMMs with 3D Capabilities

Unified Human-Scene Interaction via Prompted Chain-of-Contacts

Monocular 3D Object Detection with Depth from Motion

MV-JAR: Masked Voxel Jigsaw and Reconstruction for LiDAR-Based Self-Supervised Pre-Training

GenNBV: Generalizable Next-Best-View Policy for Active 3D Reconstruction

GeoMIM: Towards Better 3D Knowledge Transfer via Masked Image Modeling for Multi-view 3D Understanding

RoboGround: Robotic Manipulation with Grounded Vision-Language Priors

GENMANIP: LLM-driven Simulation for Generalizable Instruction-Following Manipulation

Rethinking the Embodied Gap in Vision-and-Language Navigation: A Holistic Study of Physical and Visual Disparities

OST-Bench: Evaluating the Capabilities of MLLMs in Online Spatio-temporal Scene Understanding

GLEAM: Learning Generalizable Exploration Policy for Active Mapping in Complex 3D Indoor Scene

VFlowOpt: A Token Pruning Framework for LMMs with Visual Information Flow-Guided Optimization