Picture for Hongwei Yi

Hongwei Yi

Max Planck Institute for Intelligent Systems, Germany

MagicDistillation: Weak-to-Strong Video Distillation for Large-Scale Portrait Few-Step Synthesis

Add code
Mar 17, 2025
Viaarxiv icon

MagicID: Hybrid Preference Optimization for ID-Consistent and Dynamic-Preserved Video Customization

Add code
Mar 16, 2025
Viaarxiv icon

MagicInfinite: Generating Infinite Talking Videos with Your Words and Voice

Add code
Mar 07, 2025
Viaarxiv icon

Magic 1-For-1: Generating One Minute Video Clips within One Minute

Add code
Feb 11, 2025
Viaarxiv icon

Real-time One-Step Diffusion-based Expressive Portrait Videos Generation

Add code
Dec 18, 2024
Figure 1 for Real-time One-Step Diffusion-based Expressive Portrait Videos Generation
Figure 2 for Real-time One-Step Diffusion-based Expressive Portrait Videos Generation
Figure 3 for Real-time One-Step Diffusion-based Expressive Portrait Videos Generation
Figure 4 for Real-time One-Step Diffusion-based Expressive Portrait Videos Generation
Viaarxiv icon

Human-Aware 3D Scene Generation with Spatially-constrained Diffusion Models

Add code
Jun 26, 2024
Viaarxiv icon

Generating Human Interaction Motions in Scenes with Text Control

Add code
Apr 16, 2024
Viaarxiv icon

DECO: Dense Estimation of 3D Human-Scene Contact In The Wild

Add code
Sep 26, 2023
Viaarxiv icon

POCO: 3D Pose and Shape Estimation with Confidence

Add code
Aug 24, 2023
Viaarxiv icon

TADA! Text to Animatable Digital Avatars

Add code
Aug 21, 2023
Viaarxiv icon