Peng Li

YuE: Scaling Open Foundation Models for Long-Form Music Generation

PanoLAM: Large Avatar Model for Gaussian Full-Head Synthesis from One-shot Unposed Image

MeshMosaic: Scaling Artist Mesh Generation via Local-to-Global Assembly

TrackingWorld: World-centric Monocular 3D Tracking of Almost All Pixels

SyncHuman: Synchronizing 2D and 3D Generative Models for Single-view Human Reconstruction

PSHuman: Photorealistic Single-view Human Reconstruction using Cross-Scale Diffusion

$CoCoGesture$: Towards Coherent Co-speech 3D Gesture Generation in the Wild