π Empowering Coral Reef Research with AI-Powered Annotation Tools π
An unofficial toolkit to supercharge your CoralNet workflows with cutting-edge computer vision
π Project Stats
[](https://pypi.org/project/CoralNet-Toolbox)
[](https://pypi.python.org/pypi/CoralNet-Toolbox)
[](https://pepy.tech/project/coralnet-toolbox)
[](https://pypi.org/project/CoralNet-Toolbox)
[](https://pypi.org/project/CoralNet-Toolbox)
[](https://pypi.org/project/CoralNet-Toolbox)
[](https://pypi.org/project/CoralNet-Toolbox)
| π― **Smart Annotation** | π€ **AI-Powered** | π **Complete Pipeline** |
|:---:|:---:|:---:|
| Create patches, rectangles, and polygons with intelligent assistance | Leverage SAM, YOLO, and foundation models | From data collection to deployment |
| *Precision meets efficiency* | *Cutting-edge AI at your fingertips* | *End-to-end workflow automation* |
β‘ Quick Start β‘
Get up and running in seconds:
# π» Installation
pip install coralnet-toolbox
# π Launch
coralnet-toolbox
π Thatβs it! The toolbox will open and youβre ready to start annotating!
For a complete installation guide (including CUDA setup), see the Installation Documentation.
π Documentation Hub
| π Guide | π― Purpose | π Link |
|:---:|:---:|:---:|
| **Overview** | Get the big picture | [π Read More](https://jordan-pierce.github.io/CoralNet-Toolbox/overview) |
| **Installation** | Detailed setup instructions | [βοΈ Setup Guide](https://jordan-pierce.github.io/CoralNet-Toolbox/installation) |
| **Usage** | Learn the tools | [π οΈ User Manual](https://jordan-pierce.github.io/CoralNet-Toolbox/usage) |
| **Hot Keys** | Keyboard shortcuts | [β¨οΈ Shortcuts](https://jordan-pierce.github.io/CoralNet-Toolbox/hot-keys) |
| **Classification** | Community tutorial | [π§ AI Tutorial](https://jordan-pierce.github.io/CoralNet-Toolbox/classify) |
π₯ Video Demonstrations
πΊ Watch the Complete Tutorial Series
π¬ Complete playlist covering all major features and workflows
π€ AI Model Arsenal
The toolbox integrates state-of-the-art models for efficient annotation workflows:
### ποΈ **Trainable Models**
| **YOLO Family** | **Versions Available** |
|:---:|:---:|
| π¦Ύ **Legacy** | YOLOv3 β’ YOLOv4 β’ YOLOv5 |
| π **Modern** | YOLOv6 β’ YOLOv7 β’ YOLOv8 |
| β‘ **Latest** | YOLOv9 β’ YOLOv10 β’ YOLO11 β’ YOLO12 |
*Powered by the [Ultralytics](https://ultralytics.com/) ecosystem*
### π― **Segment Anything Models**
| **Model** | **Specialty** | **Use Case** |
|:---:|:---:|:---:|
| πͺΈ **SAM** | General segmentation | High-quality masks |
| π **CoralSCOP** | Coral-specific | Marine biology focus |
| β‘ **FastSAM** | Speed optimized | Real-time annotation |
| π± **MobileSAM** | Mobile-friendly | Edge deployment |
| βοΈ **EdgeSAM** | Efficient | Resource-constrained |
| π **RepViT-SAM** | Vision transformers | Advanced features |
*Powered by our [xSAM](https://Github.com/Jordan-Pierce/xSAM) integration*
### ποΈ **Visual Prompting & Foundation Models**
| **Framework** | **Models** | **Capability** |
|:---:|:---:|:---:|
| **YOLOE** | See Anything | Visual prompt detection |
| **Transformers** | Grounding DINO β’ OWLViT β’ OmDetTurbo | Zero-shot detection |
π οΈ Feature Showcase
### π **Core Annotation Tools**
|

**π― Patch Annotation** |

**π Rectangle Annotation** |

**π· Multi-Polygon Annotation** |
|:---:|:---:|:---:|
### π€ **AI-Powered Analysis**
|

**π§ Image Classification** |

**π― Object Detection** |

**π Instance Segmentation** |
|:---:|:---:|:---:|
### π¬ **Advanced Capabilities**
|

**πͺΈ Segment Anything (SAM)** |

**π Polygon Classification** |

**π Region-based Detection** |
|:---:|:---:|:---:|
### βοΈ **Editing & Processing Tools**
|

**βοΈ Cut** |

**π Combine** |

**π¨ Simplify** |
|:---:|:---:|:---:|
### π **Specialized Features**
|

**ποΈ See Anything (YOLOE)** |

**πΊοΈ LAI Classification** |
|:---:|:---:|
### π **Analysis & Exploration**
|

**π¬ Video Inference & Analytics** |

**π Data Explorer & Clustering** |
|:---:|:---:|
π§ Complete Workflow Pipeline
- π₯ CoralNet Download: Retrieve source data and annotations
- π¬ Video Processing: Extract frames from video files
- πΈ Image Import: Support for various image formats
βοΈ Annotation & Labeling
- π Manual Annotation: Intuitive point, rectangle, polygon and semantic tools
- π€ AI-Assisted: SAM, YOLO, and visual prompting models
- π Precision Editing: Cut, combine, subtract, and simplify shapes
π§ Machine Learning
- π¬ Hyperparameter Tuning: Optimize training conditions
- π Model Training: Build custom classifiers and detectors
- β‘ Model Optimization: Production-ready deployment
π Analysis & Export
- π Performance Evaluation: Comprehensive model metrics
- π― Batch Inference: Process multiple images automatically
- π₯ Video Analysis: Real-time processing with analytics
- π Multi-format Export: CoralNet, Viscore, TagLab, GeoJSON
π Roadmap
See the current tickets and planned features on the GitHub Issues Page