Siyu Zhu

Affiliations

Alibaba Cloud A.I. Lab

papers

2,364

total citations

papers (18)

Cascade Cost Volume for High-Resolution Multi-View Stereo and Stereo Matching

CVPR 2020arXiv

872

citations

CondLaneNet: A Top-To-Down Lane Detection Framework Based on Conditional Convolution

ICCV 2021arXiv

295

citations

Neural Window Fully-Connected CRFs for Monocular Depth Estimation

CVPR 2022arXiv

254

citations

Champ: Controllable and Consistent Human Image Animation with 3D Parametric Guidance

ECCV 2024arXiv

239

citations

Hallo3: Highly Dynamic and Realistic Portrait Image Animation with Video Diffusion Transformer

CVPR 2025arXiv

citations

FloorPlanCAD: A Large-Scale CAD Drawing Dataset for Panoptic Symbol Spotting

ICCV 2021arXiv

citations

Learning Camera Localization via Dense Scene Matching

CVPR 2021arXiv

citations

OpenHumanVid: A Large-Scale High-Quality Dataset for Enhancing Human-Centric Video Generation

CVPR 2025arXiv

citations

Self-Supervised Human Depth Estimation From Monocular Videos

CVPR 2020arXiv

citations

AlphaPO: Reward Shape Matters for LLM Alignment

ICML 2025arXiv

citations

High-Fidelity and Transferable NeRF Editing by Frequency Decomposition

ECCV 2024arXiv

citations

4D Diffusion for Dynamic Protein Structure Prediction with Reference and Motion Guidance

AAAI 2025arXiv

citations

Dirichlet-Constrained Variational Codebook Learning for Temporally Coherent Video Face Restoration

ICCV 2025arXiv

citations

RCP: Recurrent Closest Point for Point Cloud

CVPR 2022

citations

Retaining Knowledge and Enhancing Long-Text Representations in CLIP through Dual-Teacher Distillation

CVPR 2025

citations

Siyu Zhu

Affiliations

papers (18)

Cascade Cost Volume for High-Resolution Multi-View Stereo and Stereo Matching

CondLaneNet: A Top-To-Down Lane Detection Framework Based on Conditional Convolution

Neural Window Fully-Connected CRFs for Monocular Depth Estimation

Champ: Controllable and Consistent Human Image Animation with 3D Parametric Guidance

Gaussian-Flow: 4D Reconstruction with Dynamic 3D Gaussian Particle

End-to-End Learning Local Multi-View Descriptors for 3D Point Clouds

Tora: Trajectory-oriented Diffusion Transformer for Video Generation

Hallo3: Highly Dynamic and Realistic Portrait Image Animation with Video Diffusion Transformer

FloorPlanCAD: A Large-Scale CAD Drawing Dataset for Panoptic Symbol Spotting

Learning Camera Localization via Dense Scene Matching

OpenHumanVid: A Large-Scale High-Quality Dataset for Enhancing Human-Centric Video Generation

Self-Supervised Human Depth Estimation From Monocular Videos

AlphaPO: Reward Shape Matters for LLM Alignment

High-Fidelity and Transferable NeRF Editing by Frequency Decomposition

4D Diffusion for Dynamic Protein Structure Prediction with Reference and Motion Guidance

Dirichlet-Constrained Variational Codebook Learning for Temporally Coherent Video Face Restoration

RCP: Recurrent Closest Point for Point Cloud

Retaining Knowledge and Enhancing Long-Text Representations in CLIP through Dual-Teacher Distillation

papers (18)

Cascade Cost Volume for High-Resolution Multi-View Stereo and Stereo Matching

CondLaneNet: A Top-To-Down Lane Detection Framework Based on Conditional Convolution

Neural Window Fully-Connected CRFs for Monocular Depth Estimation

Champ: Controllable and Consistent Human Image Animation with 3D Parametric Guidance

Gaussian-Flow: 4D Reconstruction with Dynamic 3D Gaussian Particle

End-to-End Learning Local Multi-View Descriptors for 3D Point Clouds

Tora: Trajectory-oriented Diffusion Transformer for Video Generation

Hallo3: Highly Dynamic and Realistic Portrait Image Animation with Video Diffusion Transformer

FloorPlanCAD: A Large-Scale CAD Drawing Dataset for Panoptic Symbol Spotting

Learning Camera Localization via Dense Scene Matching

OpenHumanVid: A Large-Scale High-Quality Dataset for Enhancing Human-Centric Video Generation

Self-Supervised Human Depth Estimation From Monocular Videos

AlphaPO: Reward Shape Matters for LLM Alignment

High-Fidelity and Transferable NeRF Editing by Frequency Decomposition

4D Diffusion for Dynamic Protein Structure Prediction with Reference and Motion Guidance

Dirichlet-Constrained Variational Codebook Learning for Temporally Coherent Video Face Restoration

RCP: Recurrent Closest Point for Point Cloud

Retaining Knowledge and Enhancing Long-Text Representations in CLIP through Dual-Teacher Distillation