Ecosyste.ms: Awesome
An open API service indexing awesome lists of open source software.
https://github.com/lasithaamarasinghe/real-time-speech-recognition
This project includes a system that can record live speech using your microphone and then transcribe it using speech recognition.
https://github.com/lasithaamarasinghe/real-time-speech-recognition
ipywidgets jupyter-notebook machine-learning pyaudio pydub python3 pytorch realtime-speech-recognition speech-to-text transformers vosk
Last synced: 8 days ago
JSON representation
This project includes a system that can record live speech using your microphone and then transcribe it using speech recognition.
- Host: GitHub
- URL: https://github.com/lasithaamarasinghe/real-time-speech-recognition
- Owner: LasithaAmarasinghe
- Created: 2024-06-08T04:38:59.000Z (5 months ago)
- Default Branch: main
- Last Pushed: 2024-06-16T06:13:08.000Z (5 months ago)
- Last Synced: 2024-06-17T22:55:45.594Z (5 months ago)
- Topics: ipywidgets, jupyter-notebook, machine-learning, pyaudio, pydub, python3, pytorch, realtime-speech-recognition, speech-to-text, transformers, vosk
- Language: Jupyter Notebook
- Homepage:
- Size: 14.6 KB
- Stars: 1
- Watchers: 1
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
Awesome Lists containing this project
README
# Real-Time-Speech-Recognition
## Overview
* This project includes a system that can record live speech using the microphone and then transcribe it using speech recognition.
* This can be used to automatically record and transcribe meetings, lectures, and other events.
* This repository contains all the codes and resources of this project.
## Steps
* Creating Jupyter widgets to record audio and stop recording
* Using pyaudio to record microphone audio
* Creating a speech recognition system using vosk
* Adding punctuation to the text transcript using recasepunc
## Code
You can find the code for this project here.
* [microphone.ipynb](https://github.com/LasithaAmarasinghe/Real-Time-Speech-Recognition/blob/main/microphone.ipynb).
## Technologies/Tools
* Jupyter Notebook / JupyterLab
* Python 3.10.12
* Pytorch `pip install torch -f https://download.pytorch.org/whl/torch_stable.html`
* Python packages
* vosk `pip install vosk`
* pydub `pip install pydub`
* transformers `pip install transformers`
* pyaudio `pip install pyaudio`
* ipywidgets `pip install ipywidgets`
## Installation Guidelines
### Vosk
You need to download a model file to run vosk properly. This automatically downloads when you run this code:
```
from vosk import Model
Model(model_name="vosk-model-small-en-us-0.15")
```
The full vosk model is large (1GB+). If you want to use it, just specify `vosk-model-en-us-0.22` as the model name.
If the models don't automatically download, you can find them [here](https://alphacephei.com/vosk/models).
### Punctuation
By default, vosk outputs text with no punctuation. To add in punctuation, we need a different model. To get this, follow these steps:
* Download the model [here](https://alphacephei.com/vosk/models/vosk-recasepunc-en-0.22.zip) - caution: 1GB+ in size.
* Extract the zip file into the same directory as your code.
### Pyaudio
Pyaudio can be a little tricky to install, since it depends on system packages. Check the [homepage](http://people.csail.mit.edu/hubert/pyaudio/) for specific instructions for each OS.
You also want to figure out the right device to record from. Run this code to find the index of your microphone:
```
# Find audio device index
import pyaudio
p = pyaudio.PyAudio()
for i in range(p.get_device_count()):
print(p.get_device_info_by_index(i))
p.terminate()
```
## Data
All audio will come from the microphone.