This repository contains a lightweight Python script designed to classify spoken words—specifically "yes" and "no". The project serves as a practical introduction to Natural Language Processing (NLP) using the popular scikit-learn library.
Before running the scripts, ensure you have Python installed. You can install all necessary dependencies using the provided requirements.txt:
pip install -r requirements.txt
The dataset is not included in this repository due to its size. Please follow these steps to set up the environment:
- Download: Get the dataset from TensorFlow Speech Commands.
- Extract: Uncompress the downloaded archive.
- Structure: Place the files in a
data/directory. Your folder structure should look like this:
| Path | Description |
|---|---|
data/speech_commands_v0.02/ |
Root folder for the speech data |
data/speech_commands_v0.02/yes/ |
Audio samples for "yes" |
data/speech_commands_v0.02/no/ |
Audio samples for "no" |
For a deeper dive into the methodology, model performance, and technical details, please refer to the: