ultralytics/ultralytics
Ultralytics YOLO27, YOLO26, YOLO11, YOLOv8 — object detection, instance segmentation, semantic segmentation, image classification, pose estimation, object tracking
A star delta has been computed for 61,707 of 62,640 tracked repositories in this period, and only those can be ranked in it.
A repository missing here has not been measured far enough back yet. That does not mean it did not move.
Star counts measured between 00:35 and 02:30 (BKK).
Ultralytics YOLO27, YOLO26, YOLO11, YOLOv8 — object detection, instance segmentation, semantic segmentation, image classification, pose estimation, object tracking
X-AnyLabeling: A lightweight, efficient, and unified cross-platform desktop application for annotating text, image, video, point cloud and multimodal data, combining versatile built-in tools with state-of-the-art AI models and flexible multi-format export.
LibreYOLO is a MIT licensed open source computer vision library
SOMA: Unifying Parametric Human Body Models
Markerless kinematics with any cameras — From 2D Pose estimation to 3D OpenSim motion
[TMLR 26] EdgeCrafter: Compact ViTs for Edge Dense Prediction via Task-Specialized Distillation
High-performance Ultralytics YOLO inference in Rust with ONNX Runtime, GPU backends, CLI, and WebGPU/WASM.
Ultralytics YOLO iOS app and Swift package for real-time Core ML inference across major computer vision tasks.
Real-time 3D full-body reconstruction from a single camera, Multiperson BVH output, Pure C++ runtime, ONNX + ggml, 70-joint skeleton with hands.
YOLO Vision Studio: a free, open-source Streamlit app for detection, segmentation, pose, text prompts (YOLO World, YOLOE) and tracking with honest counts. Built on Ultralytics YOLO26.
A library for differentiable robotics on manifolds.
Deep learned, NVIDIA-accelerated 3D object pose estimation
The collection of pre-trained, state-of-the-art AI models for ailia SDK
Official implementation of DeepLabCut: Markerless pose estimation of user-defined features with deep learning for all animals incl. humans
GenZ-ICP: SOTA robust LiDAR odometry (IEEE RA-L 2025)
LightGlue: Local Feature Matching at Light Speed (ICCV 2023)
That's all 16 repositories in this period.