WorldPlay: Interactive World Modeling with Real-Time Latency and Geometric Consistency

Code for the SIGGRAPH Asia 2025 paper Gradient-Weighted Feature Back-Projection: A Fast Alternative to Feature Distillation in 3D Gaussian Splatting

[SIGGRAPH Asia 2025] Designing and Fabricating Color BRDFs with Differentiable Wave Optics

[NeurIPS 2025] PhysCtrl: Generative Physics for Controllable and Physics-Grounded Video Generation

Official repository of the "Clustered Error Correction with Grouped 4D Gaussian Splatting" (SIGGRAPH Asia 2025)

[SIGGRAPH-ASIA'25] The code of paper "Lifted Surfacing of Generalized Sweep Volumes"

Genetic Art in CUDA, C++, and OpenGL; great fractal and organic modes, unlike Karl Sims' original 1995 Siggraph paper

(Siggraph Asia 2025) Code of "LayerPeeler: Autoregressive Peeling for Layer-wise Image Vectorization"

UCAgents: Unidirectional Convergence for Visual Evidence Anchored Multi-Agent Medical Decision-Making

Official Implementation of Particulate: Feed-Forward 3D Object Articulation

Code for the paper: "Modular Neural Image Signal Processing". A modular neural ISP with interpretable stages, multi-style rendering, cross-camera generalization, and post-editable re-rendering with…

"AnyTool: Universal Tool-Use Layer for AI Agents"

Foundation Models and Data for Human-Human and Human-AI interactions.

SIGGRAPH 2025 "Practical Inverse Rendering Of Textured And Translucent Appearance"

[SIGGRAPH Asia 2025] Official github repo of SeqTex, an end-to-end 3D texture generation method using video diffusion priors.

[CVPR 2025] High-Fidelity Relightable Monocular Portrait Animation with Lighting-Controllable Video Diffusion Model

A Cross-Platform Backend for High-Performance Sparse Convolutions

[arXiv 2025] Official implementation of BiCo: Composing Concepts from Images and Videos via Concept-prompt Binding

Visionary: The World Model Carrier Built on WebGPU-Powered Gaussian Splatting Platform

[SIGGRAPH Asia 2025] WorldExplorer: Towards Generating Fully-Explorable 3D Scenes

Training code for "Radiance Meshes for Volumetric Reconstruction".

"Paper2Slides: From Paper to Presentation in One Click"

Reference code for the paper: "Raw-JPEG Adapter: Efficient Raw Image Compression with JPEG." Learnable, invertible raw-to-JPEG pre-processing with parameters stored in the JPEG COM (<64 KB), enabli…

GECO: Generative Image-to-3D within a SECOnd