Starred repositories
[arXiv] Jeveration: System One Models Now Generate. https://arxiv.org/abs/2611.24738
Public code release for the paper "HyperThink: Text-to-Parameter Hypernetworks for Efficient Reasoning"
The open-source Claude Motion & CapCut alternative. Turn a reference video into editable motion graphics with your brand kit. Export MP4s for YouTube, YouTube Shorts, Instagram Reels, and TikTok. A…
Proactive agent that grows with you
A minimalist and powerful harness with one tool: a persistent Python REPL.
[ICLR2026] FastAvatar: Towards Unified and Fast 3D Avatar Reconstruction with Large Gaussian Reconstruction Transformers
[CVPR '26] FlexAvatar: Learning Complete 3D Head Avatars with Partial Supervision
[ECCV 2026] PointSplat: Compact Gaussian Splatting via Human-Centric Prediction
Official repo for PhysAvatar: Learning the Physics of Dressed 3D Avatars from Visual Observations, ECCV 2024
Official repository for CVPR 2024 highlight paper 4D-DRESS: A 4D Dataset of Real-world Human Clothing with Semantic Annotations.
EditaLive! Unified Character Video Editing for Live Streaming
MobilePortrait ported to MLX for real-time face animation on Apple Silicon (M3 Max); incl. Core AI / ANE conversion experiments
[ECCV 2026 Oral] DreamID-V: Bridging the Image-to-Video Gap for High-Fidelity Face Swapping via Diffusion Transformer
Unlimited-length talking video generation that supports image-to-video and video-to-video generation
Real-time, long-horizon audio-driven avatar generation via decoupled parallel training.
Official repo for "PRISM: Prior-Rich Single-Image Animatable Avatar via Synergistic Multi-Source Supervision"
Official implementation of DiGS-Avatar: Single-Image Animatable 3D Human Reconstruction via UV-Space Diffusion
HumanNOVA: Photorealistic, Universal and Rapid 3D Human Avatar Modeling from a Single Image (CVPR 2026 Highlight)
LHM++: An Efficient Large Human Reconstruction Model for Pose-free Images to 3D
[CVPR'25] ChatGarment: Garment Estimation, Generation and Editing via Large Language Models
ReMu: Reconstruct Multi-layer 3D Clothed Humans from Images