Structured Chain-of-Thought

An efficient and research-friendly Gaussian Splatting framework described in the CVPR'26 paper "Faster-GS: Analyzing and Improving Gaussian Splatting Optimization"

Official implementation of "Speeding Up the Learning of 3D Gaussians with Much Shorter Gaussian Lists" (CVPR 2026)

Open-source code for the first-place solution of the [SIGGRAPH Asia 2025 3DGS Challenge](https://gaplab.cuhk.edu.cn/projects/gsRaceSIGA2025/index.html).

A machine learning-based video super resolution and frame interpolation framework. Est. Hack the Valley II, 2018.

WebGL effects made easy.

[Official Code] Pixel3DMM: Versatile Screen-Space Priors for Single-Image 3D Face Reconstruction

An emulation engine to improve Warcraft III modding

Tiny AutoEncoder for Hunyuan Video (and other video models)

Official code for "LagerNVS Latent Geometry for Fully Neural Real-time Novel View Synthesis" (CVPR 2026)

[CVPR26] GeoMotion: Rethinking Motion Segmentation via Latent 4D Geometry

WorldTree: Towards 4D Dynamic Worlds from Monocular Video using Tree-Chains (ICLR2026)

[CVPR 2026] Mesh4D: 4D Mesh Reconstruction and Tracking from Monocular Video

[ICLR 2026] Light-X: Generative 4D Video Rendering with Camera and Illumination Control

[CVPR 2026] Official pytorch implementation of "ReDirector: Creating Any-Length Video Retakes with Rotary Camera Encoding"

[CVPR'25] SplineGS: Robust Motion-Adaptive Spline for Real-Time Dynamic 3D Gaussians from Monocular Video

Code for the project "MegaSaM: Accurate, Fast and Robust Structure and Motion from Casual Dynamic Videos"

[NeurIPS 2025] Instant4D: 4D Gaussian Splatting in Minutes

[ICLR 2026] PAGE-4D: Disentangled Pose and Geometry Estimation for VGGT-4D Perception

[CVPR 2026 Highlight🔥] MotionCrafter: Dense Geometry and Motion Reconstruction with a 4D VAE

[CVPR'2026]: MoRe: Motion-aware Feed-forward 4D Reconstruction Transformer

4DNeX: Feed-Forward 4D Generative Modeling Made Easy

[ICLR 2026] Official implementation of "SeC: Advancing Complex Video Object Segmentation via Progressive Concept Construction"

[ICCV 2025] SAM2Long: Enhancing SAM 2 for Long Video Segmentation with a Training-Free Memory Tree

Unofficial DynaDUSt3R reimplementation trained on Stereo4D (research only).