africa-ai/streamlit-tools

A Streamlit application for editing and correcting transcriptions of Kalenjin audio recordings. # Kalenjin Audio Transcription Editor A Streamlit application for editing and correcting transcriptions of Kalenjin audio recordings. ## Features - **File Management:** Use included data files or upload your own - **Audio Playback:** Listen to each audio segment directly in the browser - **Transcription Editing:** Compare and correct transcriptions - **Progress Tracking:** Monitor your editing progress with visual indicators - **History:** Keep track of all changes made to transcriptions - **Export:** Download your corrected transcriptions as CSV ## Getting Started ### Prerequisites - Python 3.7 or higher - Required Python packages: - streamlit - pandas - librosa - soundfile - numpy ### Installation 1. Clone this repository: ``` git clone cd streamlit-tools ``` 2. Install required packages: ``` pip install -r requirements.txt ``` 3. Run the application: ``` streamlit run main.py ``` ### Using the Application 1. Choose your data source: - Use the provided files in the data folder - Upload your own transcriptions CSV and audio chunks ZIP 2. Navigate through transcriptions using the Previous and Next buttons 3. Listen to each audio segment and correct the transcription as needed 4. Save your corrections and they will be automatically stored 5. Download the corrected transcriptions when finished ## Data Organization 1. The transcriptions CSV should contain at least two columns: - `chunk_filename`: Path to the audio file - `transcription`: The text transcription for the audio file 2. The audio chunks ZIP should contain WAV files with names matching those in the transcriptions file ## License This project is licensed under the MIT License - see the LICENSE file for details. ## Acknowledgments - Streamlit for the excellent web app framework - Contributors to the open-source libraries used in this project