PyTorch3D is FAIR's library of reusable components for deep learning with 3D data

Code release for NeRF (Neural Radiance Fields)

Depth Anything 3

MapAnything: Universal Feed-Forward Metric 3D Reconstruction

Generalizable Perception Stack for all things 3D, 4D & Scene Understanding

Official code for "FeatUp: A Model-Agnostic Frameworkfor Features at Any Resolution" ICLR 2024

Code of π^3: Permutation-Equivariant Visual Geometry Learning

A geometry-shader-based, global CUDA sorted high-performance 3D Gaussian Splatting rasterizer. Can achieve a 5-10x speedup in rendering compared to the vanialla diff-gaussian-rasterization.

[CVPR 2025 Best Paper Award] VGGT: Visual Geometry Grounded Transformer

[ICCV 2025] A simple training-free approach adapting DUSt3R for dynamic scenes.

DELTA: Dense Efficient Long-range 3D Tracking for Any video (ICLR 2025)

[CVPR 2024 Highlight] Official PyTorch implementation of SpatialTracker: Tracking Any 2D Pixels in 3D Space

[CVPR 2025] MIDI: Multi-Instance Diffusion for Single Image to 3D Scene Generation

[ICCV 2025] Official impl. of "MV-Adapter: Multi-view Consistent Image Generation Made Easy"

Original reference implementation of "3D Gaussian Splatting for Real-Time Radiance Field Rendering"

[NeurIPS 2023] Official implementation of the paper "Motion-X: A Large-scale 3D Expressive Whole-body Human Motion Dataset"

[AAAI 2026] VisionReward: Fine-Grained Multi-Dimensional Human Preference Learning for Image and Video Generation

✨✨Latest Advances on Multimodal Large Language Models

Pytorch implementation of same-family gaussian mixture models with guardrails. Features separable parameter optimization and singularity mitigation

TorchCFM: a Conditional Flow Matching library

Jupyter notebook with Pytorch implementation of Neural Ordinary Differential Equations

A generative world for general-purpose robotics & embodied AI learning.

A data generation pipeline for creating semi-realistic synthetic multi-object videos with rich annotations such as instance segmentation masks, depth maps, and optical flow.

Bullet Physics SDK: real-time collision detection and multi-physics simulation for VR, games, visual effects, robotics, machine learning etc.

Interactive visualizations of the geometric intuition behind diffusion models.

A PyTorch library for implementing flow matching algorithms, featuring continuous and discrete flow matching implementations. It includes practical examples for both text and image modalities.

Efficient Track Anything

[Neurips 2023 & TPAMI] T2I-CompBench (++) for Compositional Text-to-image Generation Evaluation