README.md

# Dataset Directory Structure

<!--
Project: Video Classification Dataset
Author: Molla Samser
Designer & Tester: Rima Khatun
Website: https://rskworld.in
Email: help@rskworld.in, support@rskworld.in
Phone: +91 93305 39277
Organization: RSK World
-->

## Directory Structure

```
data/
├── train/ # Training videos (70% of dataset)
│ ├── action/
│ ├── comedy/
│ ├── drama/
│ ├── sports/
│ ├── documentary/
│ ├── news/
│ ├── music/
│ └── education/
├── test/ # Test videos (20% of dataset)
│ ├── action/
│ ├── comedy/
│ ├── drama/
│ ├── sports/
│ ├── documentary/
│ ├── news/
│ ├── music/
│ └── education/
└── validation/ # Validation videos (10% of dataset)
├── action/
├── comedy/
├── drama/
├── sports/
├── documentary/
├── news/
├── music/
└── education/
```

## Adding Videos

1. Place your video files in the appropriate category folder
2. Supported formats: MP4, MOV, AVI, MKV
3. Use the `organize_dataset.py` script to automatically split videos into train/test/validation sets

## Video Requirements

- Format: MP4, MOV, AVI, or MKV
- Recommended resolution: 224x224 or higher
- Recommended duration: 5-10 seconds per clip
- Frame rate: 24-30 FPS

## Example Usage

```bash
# Add videos to raw_videos directory organized by category
raw_videos/
├── action/
│ ├── video1.mp4
│ └── video2.mp4
└── comedy/
├── video1.mp4
└── video2.mp4

# Then organize into train/test/validation
python scripts/organize_dataset.py --input raw_videos --output data
```