Stars
Reproducible unsigned iPadOS builds of Zed PR #52921 using macOS GitHub Actions
Native Android port of the Zed code editor, gpui rendering through Vulkan, integrated Termux for LSPs and tooling, remote SSH support. Runs as an APK on tablets and Android desktop modes.
[ECCV 2026] Wide-angle pose refinement for unseen objects from RGB-D, camera intrinsics, a metric mesh, and a mask or box.
PointWorld: Scaling 3D World Models for In-The-Wild Robotic Manipulation
Turn your Panasonic Lumix camera into a Fujifilm GFX, Leica, ARRI, and more using precise ACES inverse engineering
Unofficial WB-dependent HNCS color and tone reimplementation for Adobe Lightroom RAW workflows.
[NeurIPS 2026 Spotlight] MoSE3: Learning World-Space SE(3) at Every Pixel
This repository contains utilities for loading data from Meta's datasets with hand pose annotations, computing hand tracking metrics, and visualising results of a hand tracking method.
CVPR'24: MS-MANO: Enabling Hand Pose Tracking with Biomechanical Constraints
CVPR'24: MS-MANO: Enabling Hand Pose Tracking with Biomechanical Constraints
Perspective Crop Layers in PyTorch
Wearable battery powered IMU to Bluetooth module.
Physical AI training data calculated, visualized, transmitted and recorded.
[CVPR 2024] MANUS: Markerless Grasp Capture using Articulated 3D Gaussians
Mesa is an open-source Python library for agent-based modeling, ideal for simulating complex systems and exploring emergent behaviors.
Official implementation of WorldinWorld
Code for SM4RT: Learning Structured Motion Geometry for 4D Reconstruction
[CVPR 2024 Highlight] Official PyTorch implementation of SpatialTracker: Tracking Any 2D Pixels in 3D Space
ViPE: Video Pose Engine for Geometric 3D Perception
BVB: Benchmarking Agentic Video Understanding via Programmatic Reconstruction in Blender
An open-source, GPU-accelerated physics simulation engine built upon NVIDIA Warp, specifically targeting roboticists and simulation researchers.
WeMM-Embedding is a family of universal multimodal embedding models by the WeChat Vision Team at Tencent, supporting multimodal understanding and retrieval.
[SIGGRAPH Asia 2026] 4DAnyone: Create Anyone in 4D from a Casual Monocular Video
[ICLR2025, ICML2025, NeurIPS2025 Spotlight] Quantized Attention achieves speedup of 2-5x compared to FlashAttention, without losing end-to-end metrics across language, image, and video models.