/awesome-segment-anything-extensions

Segment-anything related awesome extensions/projects/repos.

Awesome-segment-anything-extensions Awesome Awesome Anything

Segment-anything related awesome extensions/projects/repos/applications.

Welcome to starts ⭐ & comments 💹 & sharing 😀 !!

- 2023.4.9: Recent works/projects (from 2023) 
- welcome to add if any information misses. 😎

Recent Works

  • Segment Anything Model (SAM) official Repo

The repository provides code for running inference with the SegmentAnything Model (SAM), links for downloading the trained model checkpoints, and example notebooks that show how to use the model.

image

  • Grounded-Segment-Anything (Combination of Ground-DINO and Segment-Anything) Repo

image

  • Segment-Any-Anomaly + (SAM + hybrid prompt regularization for zero-shot anomaly detection)ArXiV

image

image

  • Segment Any Point Cloud Sequences (SAM for self-supervised learning on large-scale 3D point clouds) Repo ArXiv Page

    image

  • Segment-Any-Anomaly (SAM + GroudingDINO for anomaly detection)Repo

image

  • Segment and Track Anything (SAM-Track) Repo ArXiv

image

  • Inpaint/Remove/Fill-Anything (image inpainting meets sam) Repo

image

  • SEEM: Segment Everything Everywhere All at Once Repo

image

  • Roboflow Annotate Smart Polygon is a free, automated image annotation tool powered by Segment-Anything Blog, App
roboflow-sam.mp4
  • Image Editing Anything (SAM + CLIP +StableDiffusion) Repo

image

  • Semantic Segment Anything (SAM for semantic seg) Repo

image

  • segment-anything-with-clip (SAM + CLIP) Repo

image

  • magic-copy (use SAM for foreground mask segmentation) Repo

image

  • EditAnything (SAM for image generation) Repo

image

  • Prompt-Segment-Anything (use SAM for instance segmentation) Repo

image

  • RefSAM (use SAM for refer image segmentation) Repo

  • sam-mmrotate(use SAM for generating rotated bounding boxes with MMRotate) Repo

image

  • Fast-SAM (fast segment-anything) repo

image

  • Mobile-SAM (faster segment-anything) repo

image

  • Disappear(use SAM for foreground removal) Repo

image

  • Segment-anything-with-image-captioning(sam + image caption) Repo

  • Open-vocabulary-Segment-Anything (OWL-ViT of Google + Segment Anything of Meta) Repo

image

  • semantic drawing with Segment Anything Model (SAM for mask drawing/editing) Repo

image

  • Annotation-anything-pipeline (GPT+SAM) Repo image

  • MOTSAM (MOT+SAM->MOTS)Repo

  • SAMCOD (SAM for Camouflaged Object Detection) Paper Repo

  • Segment Anything Model (SAM) in Napari (SAM + Napari) Repo

image

  • inpainting-anything (SAM + inpainting) Repo

image

  • segment-anything-eo (Earth observation tools for Meta AI Segment Anything) Repo

image

  • Segment Anything for Stable Diffusion Webui (helping stable diffusion webui users to use segment anything to do stable diffusion inpainting) Repo

image

image

  • segment-anything-and-name-it (Visual GPT + SAM) Repo

image

  • AnyLabeling = LabelImg + Labelme + Improved UI + Auto-labeling Repo

image

  • Caption-anything (Segment Anything + Visual Captioning + ChatGPT) Repo

image

  • Segment-anything-u-specify (using sam + clip to segment anything u specify) Repo

image

  • OCR-SAM (SAM for Text, Text Removal and Text Inpainting) Repo

image

image

  • finetune-anything (fine-tuning of SAM ) Repo

image

  • grounded-segment-anything-colab Repo

image

  • SAM Medical Imaging Repo

image

  • Segment Anything Prompt (SAM + GroundDINO + Stable Diffusion for foreground removal and background inpainting) Repo

image

  • Transform Image Into Unique Paragraph (ChatGPT+SAM+BLIP) Repo

image

  • LIME-SAM (SAM + Explainable Artificial Intelligence) Repo

image

  • sam-with-mmdet (SAM+MMDetection) Repo

image

  • open-tracking (sam + tracking) Repo

image

  • 3D-Box via Segment Anything (sam for 3D Detection) Repo

image

  • Anything-3D (Segment-Anything + 3D. Let's lift the anything to 3D) Repo

image

image

  • Transfer-Any-Style (SAM for style transfer) Repo

image

  • Paint-anything (sam for drawing) Repo

image

  • track-anything (VOS + sam)Repo

image

  • SAM-medical (SAM for medocal image segmentation)Repo

  • relate-anything (relation empowered via SAM) Repo

image

  • Count Anything (SAM for counting) Repo

image

  • ShowAnything (SAM + SD for editing) Repo

image

  • MedSAM (SAM for medicine, training tutorial) Repo

  • Personalize-SAM (Personalize Segment Anything with 1 Shot in 10 Seconds) Repo image

  • Segment Any Moving Object with Moving Ego Vehicle (SAM for multi-object segmentation) arXiv image

  • Customized Segment Anything Model for Medical Image Segmentation (SAM+LoRA fot Medical Segmentation) ArXiv

image

  • Instruct to Act (SAM+OpenCLIP for robotics application) Repo

image

  • finetune-anything (SAM finetuning) Repo

image

  • SAMScore (semantic Structural Similarity Metric for Image Translation Evaluation) ArXiv

image

  • Detect-AnyShadow (SAM for shadow det) Repo

image

  • SAD: Segment Any RGBD (SAM+Depth / SAM for RGBD) Repo

image

  • SAM-HQ: Segment Anything in High Quality (High Quality Mask with SAM) Repo

image

  • Calib-anything (SAM for calibration optimization) Repo

image

Relevant Repos

  • Towards Segmenting Anything That Moves Repo

image

  • SegGPT: Segmenting Everything In Context Repo

image

image

  • Stable Diffusion (High-Resolution Image Synthesis with Latent Diffusion Models) Repo

image

  • ControlNet (Adding Conditional Control to Text-to-Image Diffusion Models.) Repo Paper

image

  • image2text powered by CLIP and BLIP Repo