little_sophia_brain/README.md

# Little Sophia's Brain

A modular Python script designed to run on Little Sophia's Radxa ZERO 3W. It handles higher-level input processing outside the abilities of the ESP32 central controller.

- Face Detection
- Voice-to-Text (V2T)

---

## Installation

1. Clone this repository to your Radxa ZERO 3W.
2. Download and place the Vosk model in the install directory:
   - [vosk-model-small-en-us-0.15](https://alphacephei.com/vosk/models)
3. (Jake needs to add the rest of the prerequisites here — e.g., Python packages, environment setup, etc.)


## TODO:
Tested with usb mic and camera only (Logitech C920)

## File Overview

| File              | Description                                                                 |
|------------------|------------------------------------------------------------------------------|
| `RetinaFace.rknn` | RKNN model for face detection and video streaming.                          |
| `main.py`         | Core script. Integrates face detection and voice-to-text into Flask stream. |
| `realtime.py`     | Voice-to-text module (Vosk). For testing separately.                        |