# Commands Dataset

A collection of English voice commands focused on two main categories: general commands and food ordering interactions. The dataset includes audio recordings with corresponding transcriptions and detailed contextual metadata.

## Dataset Description

This dataset contains 91 voice commands recorded in various settings (office, restaurant, coffee shop, fast food) with consistent audio quality and moderate noise levels. Each command is recorded with international English accent and includes detailed contextual information.

### Key Features

- **Language**: English (en)
- **Accent**: International
- **Total Commands**: 91
- **Command Types**: 2 categories (Command, Order_Food)
- **Venue Types**: 4 categories (office, restaurant, coffee_shop, fast_food)
- **Noise Level**: Moderate
- **Audio Format**: M4A

### Command Categories

1. **General Commands** (office setting):
   - Movement commands (e.g., "Go left", "Turn right", "Back up")
   - Speed commands (e.g., "Faster", "Slow down")
   - Action commands (e.g., "Sit down", "Stop", "Come here")

2. **Food Ordering** (various venues):
   - Restaurant orders
   - Coffee shop orders
   - Fast food orders
   - Beverage orders

### Data Format

Each entry contains:
```json
{
  "audioFile": "string",
  "commandText": "string",
  "commandType": "string",
  "venueType": "string",
  "venue_name": "string",
  "language": "string",
  "accent": "string",
  "noise_level": "string",
  "backgroundNoiseType": "string"
}
```

## Usage

This dataset is particularly useful for:
- Voice command recognition systems
- Restaurant ordering automation
- Virtual assistants
- Natural language processing
- Speech recognition model training
- Command intent classification

### Loading the Dataset

```python
from datasets import load_dataset

dataset = load_dataset("maxF6YsK/commands")
```

## Dataset Statistics

### Command Distribution
- Total commands: 91
- Command types: 2 (Command, Order_Food)
- Venue types: 4 (office, restaurant, coffee_shop, fast_food)

### Audio Characteristics
- Format: M4A
- Noise Level: Moderate
- Background Noise Types: none, traffic, conversation
- Accent: International English

### Common Patterns
- Polite forms ("please", "excuse me")
- Direct commands
- Food/beverage ordering variations
- Movement and position instructions

## Data Collection Environment

- **Language**: English
- **Accent**: International
- **Noise Level**: Moderate
- **Recording Quality**: Consistent across samples
- **Background Conditions**: Controlled with minimal variation