segment-anything-2 开源项目推荐（28 个）

vietanhdev /anylabeling

在YOLO、Segment Anything（Sam+ SAM 2/2.1+ SAM 3）、ALESAM 3的AI支持下，轻松实现AI辅助数据标签！

yatengLG /ISAT_with_segment_anything

带Sam（segment anything models）的标签工具，支持Sam、Sam2、Sam3、sam-hq、Mobile Sam EdgeSam等。交互式半自动图像标记工具

ImprintLab /Medical-SAM2

Medical SAM 2: Segment 3D Medical Images Via Segment Anything Model 2

SegmentationBLWX /sssegmentation

SSSegmentation：一个基于 PyTorch 的开源有监督语义分割工具箱。

zgcr /SimpleAICV_pytorch_training_examples

SimpleAICV：pytorch 训练示例。

ibaiGorordo /ONNX-SAM2-Segment-Anything

Python scripts for the Segment Anythin 2 (SAM2) model in ONNX

autodistill /autodistill-grounded-sam-2

Use Segment Anything 2, grounded with Florence-2, to auto-label data for use in training vision models.

SauravMaheshkar /samv2

CPU compatible fork of the official SAMv2 implementation aimed at more accessible and documented tutorials

jhj0517 /sam2-playground

Playground Web UI using segment-anything-2 models from the Meta.

sujanshresstha /SAM2-in-video

This repository contains code for deploying a Gradio application using the SAM2 model for video processing. The application allows users to interact with the model through a user-friendly web interface.

VVJia /SAM-Swin

This is an official implementation for "SAM-Swin: SAM-Driven Dual-Swin Transformers with Adaptive Lesion Enhancement for Laryngo-Pharyngeal Tumor Detection" - Medical Image Analysis (Major Revision)

zz990099 /sam2_cpp

SAM2 Track implementation with TensorRT & OnnxRuntime

F5-AI /SAM2-Gradio

A gradio based webui for meta segment-anything-model 2 (SAM2), both image and video are supported

bladeszasza /SOWLv2

(SegmentedOWLv2) is a powerful command-line tool for text-prompted object segmentation for video and images.

VisioSphereAI /labelvim

This is a python based standalone image annotation tool designed for tasks such as image segmentation and object detection. The tool allows users to draw bounding boxes, polygons, and freehand paths, edit and manage annotations, and export the annotated data in multiple formats including COCO, YOLO, and PASCAL VOC.

DhanvinG /CataractSAM-2

Code for "CataractSAM-2: Enhancing Transferability and Real-Time Ophthalmic Surgery Segmentation Through Automated Ground-Truth Generation"

jing-bi /lang-ground

Use natural language to ground relevant things.

eldan1z /video-segmentation-sam2

Video Vehicle Segmentation project made with META's Segment Anything Model 2 (SAM2)

Kazuhito00 /SAM2-Colaboratory-Sample

Colaboratory上でSAM2をお試しするサンプル

jutsuhq /edgetam-onnx

Click once, track the object through video: EdgeTAM and SAM 2.1 tiny in Rust on ONNX Runtime's CUDA provider. ~42 ms a frame, mean mask IoU 0.97–0.998 vs PyTorch, no Python at run time. Windows and Linux.