Effortless AI-assisted data labeling with AI support from YOLO, Segment Anything (SAM+SAM2), MobileSAM!!

Labeling tool with SAM(segment anything model),supports SAM, SAM2, sam-hq, MobileSAM EdgeSAM etc.交互式半自动图像标注工具

Tailor是一款视频智能裁剪、视频生成和视频优化的视频剪辑工具。目前的目标是通过人工智能技术减少视频剪辑的繁琐操作，让普通人也能简单实现专业剪辑人的水准！长远目标是让视频剪辑实现真正的AIGC！

Video-Inpaint-Anything: This is the inference code for our paper CoCoCo: Improving Text-Guided Video Inpainting for Better Consistency, Controllability and Compatibility.

Grounded Tracking for Streaming Videos

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

Ultralytics VSCode snippets plugin to provide quick examples and templates of boilerplate code to accelerate your code development and learning.

Simple Video Summarization using Text-to-Segment Anything (Florence2 + SAM2) This project provides a video processing tool that utilizes advanced AI models, specifically Florence2 and SAM2, to detect and segment specific objects or activities in a video based on textual descriptions.

Image segmentation application that utilizes the SAM2 (Segment Anything Model) via API to perform object detection and segmentation on uploaded images.

This repository demonstrates the use of **SAM2 (Segment Anything Model 2)** to automatically generate object masks for images. SAM2 efficiently processes prompts to generate masks by sampling over the entire image and predicting multiple masks from single-point input prompts.

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

Colaboratory上でSAM2をお試しするサンプル

A script that utilises Facebook's SAM-2 model to add segmentation mask, bounding box, and rotation angle annotations to the MIDV500 and MIDV2019 datasets.

This repository contains an implementation of object detection using the Segment Anything Model 2 (SAM2) for product images. The current implementation focuses on detecting 'can_chowder' objects as a proof of concept.

Foundation Models Hands-On session for Cyber2A workshop

Research mainly focuses on developing circuits identifying algorithms that produce swift recognitions for industrial applications. This is done by feeding the bounding box prompt conducted by YOLOX to initiate interactive training on SAM2. Its improvement over SAM and SlimSAM will be recorded and analyzed.

SAM-2 and classification model, training on ultrasonic images dataset for predicting and segmenting breast tumors