This repository contains a simple Speech Recognition App developed using Streamlit, which allows users to transcribe spoken audio into text using various speech recognition APIs.
The Speech Recognition App is built using Python and Streamlit, utilizing the speech_recognition
library to interface with different speech recognition APIs. The app provides the following functionalities:
- User can select a speech recognition API from Google, Whisper, and Sphinx.
- User can choose the language for transcription from a list of supported languages.
- User can start recording audio through the microphone for transcription.
- Transcribed text is displayed on the app interface.
- User can clear the transcription history.
- User can download the transcription as a text file.
- Python 3.x
- Streamlit
speech_recognition
librarywhisper
library (for Whisper ASR, if using)- Active internet connection (for API-based speech recognition)
-
Install the required libraries using the following command:
pip install streamlit speech_recognition whisper
-
Clone this repository:
git clone https://github.com/yourusername/speech-recognition-app.git cd speech-recognition-app
-
Run the Streamlit app:
streamlit run app.py
-
The app will open in your browser. Select the desired API and language, then click "Start Recording" to transcribe your speech.
Access the Speech Recognition App through this link: Speech Recognition App
- The app uses Google Cloud Speech, Whisper, and Sphinx APIs for speech recognition.
- Whisper requires additional configuration and a model file.
- The Sphinx API supports only English (en-US).
- The app uses the Streamlit session state to maintain the transcription history.
Feel free to customize the app and its features according to your requirements.