# tflite-kws

**Repository Path**: lanzehua/tflite-kws

## Basic Information

- **Project Name**: tflite-kws
- **Description**: Keyword Spotting (KWS) API wrapper for TFLite streaming models.
- **Primary Language**: Unknown
- **License**: MIT
- **Default Branch**: main
- **Homepage**: None
- **GVP Project**: No

## Statistics

- **Stars**: 0
- **Forks**: 1
- **Created**: 2021-11-09
- **Last Updated**: 2021-12-29

## Categories & Tags

**Categories**: Uncategorized

**Tags**: None

## README

# TFLiteKWS

Keyword Spotting (KWS) API wrapper for TFLite streaming models.

## Model

This project is only a API wrapper to assist audio processing with tflite models. Models are not provided here.
You need to find a pretrained model or train your own. One open source streaming KWS model in tensorflow is the [kws_streaming](https://github.com/google-research/google-research/tree/master/kws_streaming) by Google Research. You can refer to [g-kws](https://github.com/StuartIanNaylor/g-kws) as a nice setup guide, and use [Dataset-builder](https://github.com/StuartIanNaylor/Dataset-builder) to prepare your own dataset.

## Testing

Simply run `python3 mic_streaming.py -m /path/to/your/model.tflite`, then speak to your mic to test.

## API

Simple example with two keywords:
```python
gkws = TFLiteKWS(args.model, [SILENCE, NOT_KW, 'keyword1', 'keyword2'])
while True:
    keyword = gkws.process(get_next_audio_frame())
    if keyword:
        # following up actions
```

Please refer to `kws.py` for detailed API usage and tunning parameters.