Project Directory Map

This document serves as the “Atlas” for the Drowsiness Detection project. It maps every file in the source code to its purpose and its corresponding documentation page.

Source Code Architecture

The project is organized into a “Pipeline” structure: Data $\to$ Training $\to$ Inference.

graph TD
    Data[data.yaml] -->|Defines| Train[train_yolo.py]
    Train -->|Produces| Model[models/*.pt]
    Model -->|Powers| Infer[drowsiness_detection.py]
    Infer -->|Triggers| Alert[Alerts: Sound/Notification]
    Req[requirements.txt] -->|Supports| All[All Scripts]

Detailed File Index

File / Folder	Technical Role	Beginner’s Explanation	Documentation Link
data.yaml	Configuration	Tells the AI where the images are.	dataset
train_yolo.py	Trainer	The script that teaches the AI.	model-training
drowsiness_detection.py	Engine	The real-time monitoring script.	detection-system
models/	Weight Storage	Stores the “learned” AI brain (.pt files).	architecture
requirements.txt	Manifest	List of all needed software libraries.	environment
runs/	Artifacts	Stores training logs and accuracy charts.	model-training
README.md	Overview	Quick-start guide.	index

Documentation Hierarchy

The documentation is structured to take you from “Curious” to “Expert”:

1. The Big Picture $\to$ index.md
2. The Setup $\to$ configuration/environment.md
3. The AI Logic $\to$ models/architecture.md $\to$ training/dataset.md $\to$ training/model-training.md
4. The Real-time Engine $\to$ inference/detection-system.md $\to$ inference/alert-mechanism.md
5. Fine Tuning $\to$ configuration/hyperparameters.md

---