Home > Categories > Computer vision > Image Segmentation > Segment Anything Model (SAM)

Segment Anything Model (SAM)

Rating:

4.8 / 5.0

Tags

Computer Vision, Image Segmentation, Deep Learning, Open Source, Meta AI

Categories:
Computer visionMachine learning and neural networks
Creator Meta AI
Date 2023-04-05
Platforms Various
Status Active (Development)
Website segment-anything.com
Price Model Free (model)
Sections:
DL FrameworksImage AnalysisImage SegmentationVideo Analysis

Pricing Details

Free (open source). Costs may apply when using third-party services or cloud resources to run the model.

Features

Image segmentation, zero-shot segmentation, promptable segmentation, high flexibility, open-source.

Integrations

Used as a model or library in Python and other language projects. Can be integrated into various computer vision applications.

Preview

The Segment Anything Model (SAM) is a new AI model from Meta AI designed for segmenting objects in images. Unlike traditional segmentation models that require training on specific object classes, SAM can segment any object in an image, even those it hasn't seen during training, based on simple prompts like a mouse click or a bounding box. The model is designed for broad application in computer vision tasks requiring precise segmentation. SAM is open-source and available to researchers and developers.