🚀 WildDetect: AI for Wildlife Conservation

Transforming Aerial Imagery into Actionable Conservation Intelligence

---

The Mission

WildDetect is a comprehensive AI-driven ecosystem designed to solve one of the most critical challenges in modern conservation: scalable and accurate wildlife monitoring.

By automating the transition from raw aerial imagery to detailed census reports, WildDetect empowers researchers and conservationists to focus on protection and policy, rather than manual image scanning.

---

🏗️ An Integrated Ecosystem

WildDetect is built on a modular three-tier architecture that mirrors the natural workflow of a data-driven conservation project:

1. The Foundation: WilData

Ensure your data is high-quality, version-controlled, and ready for intelligence. WilData handles multi-format imports (COCO, YOLO, Label Studio), geospatial metadata extraction, and large-scale image tiling.

2. The Intelligence: WildTrain

Transform raw observations into specialized AI models. WildTrain provides a flexible framework for training state-of-the-art YOLO detectors and deep-learning classifiers, integrated with MLflow for complete experiment traceability.

3. The Impact: WildDetect

Deploy your models in the field. WildDetect orchestrates final "census campaigns," processing thousands of images to generate statistically sound population counts, density maps, and professional PDF reports.

---

🗺️ How it Works: The End-to-End Workflow

WildDetect provides a seamless pipeline from raw data to field impact.

Tip

New to the project? Explore the Interactive Script Navigator in our documentation to visually map scripts and CLI commands to each step.

graph LR
    subgraph Foundation ["🗂️ 1. WilData"]
        A["Raw Images"] --> B["Processing & Tiling"]
    end
    subgraph Intelligence ["🎓 2. WildTrain"]
        B --> C["Model Training"]
        C --> D["MLflow Registration"]
    end
    subgraph Impact ["🔍 3. WildDetect"]
        D --> E["AI Detection"]
        E --> F["Census Reports"]
    end
    
    style Foundation fill:#e3f2fd,stroke:#2196f3
    style Intelligence fill:#fff8e1,stroke:#ffc107
    style Impact fill:#e8f5e9,stroke:#4caf50

Loading

---

🚀 Quick Start

1. Installation

# Clone the repository
git clone https://github.com/fadelmamar/wildetect.git
cd wildetect

# Create virtual environment (using uv)
uv venv --python 3.11
.venv\Scripts\activate  # Windows

# Install all packages
cd wildata && uv pip install -e . && cd ..
cd wildtrain && uv pip install -e . && cd ..
uv pip install -e .

2. Run Your First Detection

# Run detection using a YAML config
wildetect detection detect -c config/detection.yaml

# Run a complete census campaign
wildetect detection census -c config/census.yaml

---

📚 Documentation Reference

• Full Documentation Home
• Interactive Script Navigator
• Installation Guide
• Model Training Tutorial
• End-to-End Workflow

---

🤝 Community & Support

• Contribute: We welcome contributions! From bug reports to code improvements, check out our GitHub Issues to see what we're working on.
• Feedback: Share your conservation use cases or model results on the GitHub Discussions.

---

Developed with ❤️ for the conservation community by Seydou Fadel M. and Allin Paul.