Awesome-Segment-Anything

Tribute to Meta AI's Segment Anything Model (SAM)

A collection of projects, papers, and source code for SAM and related studies.

Keywords: Segment Anything Model, Segment Anything, SAM, awesome

CATALOGUE

Origin of the Study 💗 Project & Toolbox 💗 Lecture & Notes 💗 Papers

1 Origin of the Study

Fundemental Models

• [SAM] Segment Anything (2023)[paper] [Project]

• [SEEM] Segment Everything Everywhere All at Once (2023)[paper][Project]

2 Project & Toolbox

• [Awesome Segment-Anything Extensions] Awesome Segment-Anything Extensions [paper][code]

• [Awesome Anything] A curated list of general AI methods for Anything [paper][code]

• [Awesome Segment Anything] Awesome Segment Anything [paper][code]

Preview	Project
	[Grounded-Segment-Anything] Grounded-Segment-Anything(DINO + SAM) [paper][code]
	[Grounded Segment Anything: From Objects to Parts] Support text prompt input(GLIP/VLPart + SAM)[paper][code]
	[Semantic-Segment-Anything] A pipeline on top of SAM to predict semantic category for each mask [paper][code]
	[GroundedSAM-zero-shot-anomaly-detection] Segment any anomaly[papaer][code]
	[EditAnything] EditAnything [paper][code]
	[sd-webui-segment-anything] extension for helping stable diffusion webui with inpainting [paper][code]
	[SALT] Segment Anything Labelling Tool [paper][code]
	[SAM-CLIP] Segment Anything CLIP [paper][code]
	[Prompt-Segment-Anything] An implementation of SAM[parper][code]
	[Count-Anything] Few-shot SAM Counting[parper][code]
	[SegGPT: Vision Foundation Models] One touch for segmentation in all images (SAM+SegGPT)[paper][code]
	[napari-segment-anything] Image viewer plugin of SAM[paper][code]
	[OCR-SAM] SAM is applied to OCR [paper][code]
	[SegDrawer] Simple static web-based mask drawer[paper][code]
	[Magic Copy] Extract and copy foreground objects using SAM[paper][code]
	[Track-Anything] Video object tracking and segmentation based on SAM[paper][code]
]	[SAM-Track] Video object tracking and segmentation based on SAM[paper][code]
	[Count Anything] Count any object[paper][code]
	[Inpaint Anything] Inpaint Anything [paper][code]
	[Transfer-Any-Style] Transfer-Any-Style[paper][code]
	[Anything-3D] Anything-3D [paper][code]
	[SEEM] Segment Everything Everywhere All at Once [paper][code]
	[Caption-Anything] Generate descriptive captions for any object [paper][code]
	[Caption-Anything] Generate descriptive captions for any object [paper][code]
	[Image.txt] Transform Image Into Unique Paragraph [paper][code]
	[3D-Box via Segment Anything] Extend the scope to 3D world [paper][code]
	[SegmentAnyRGBD] Segment rendered depth images based on SAM [paper] [code]

3 Lecture & Notes

How to | roboflow how-to-segment-anything-with-sam [blog]

4 Papers

Image Segmentation & Medical Image Segmentation

• [Zero-shot Segmentation] Segment Anything Model (SAM) for Digital Pathology: Assess Zero-shot Segmentation on Whole Slide Imaging[paper][code]

• [generic segmentation] Segment Anything Is Not Always Perfect: An Investigation of SAM on Different Real-world Applications [paper][code]

• [Medical Image segmentation] SAMM (Segment Any Medical Model): A 3D Slicer Integration to SAM [paper][code]

• [Medical Image segmentation] SAM.MD: Zero-shot medical image segmentation capabilities of the Segment Anything Model [paper] [code]

• [Camouflaged Object Segmentation] SAM Struggles in Concealed Scenes -- Empirical Study on "Segment Anything"[paper][code]

• [Brain Extraction] Brain Extraction comparing Segment Anything Model (SAM) and FSL Brain Extraction Tool[paper][code]

Object Detection & Tracking

• [Camouflaged Object Detection] Can SAM Segment Anything? When SAM Meets Camouflaged Object Detection[paper][code]

• [Multi-Object Tracking] CO-MOT[paper][code]

Explaning and Supporting Basic Models

• [CLIP] CLIP Surgery for Better Explainability with Enhancement in Open-Vocabulary Tasks [paper][code]

Arxiv-daily-update

Click here to check the daily-updated paper list!