Abdulrahaman-A-Musa/hausa-audio-transcriber

Hausa audio transcription with English translation and speaker identification for survey interviews # 🎤 Hausa Audio Transcriber A powerful Streamlit web application for transcribing Hausa audio recordings with automatic English translation and speaker identification. ## 🌟 Features - **🎯 Automatic Hausa Transcription** - Uses Google Speech Recognition API - **🌐 English Translation** - Real-time translation of transcribed text - **👥 Speaker Identification** - Automatically separates Interviewer questions from Respondent answers - **⏱️ Timestamped Output** - Track when each segment was spoken - **📦 Batch Processing** - Upload and process up to 10 audio files at once - **📥 CSV Export** - Download results in structured CSV format - **🎵 Multiple Formats** - Supports WAV, MP3, M4A, AMR, AAC, 3GP, and more ## 🚀 Live Demo Try the app here ## 📋 Use Cases Perfect for: - Survey interviews transcription - Research data collection - Healthcare surveys (mortality, reproductive health) - Field research documentation - Quality control and verification ## 🛠️ Technologies Used - **Python 3.11+** - **Streamlit** - Web framework - **Google Speech Recognition** - Transcription engine - **Google Translate** - Translation service - **FFmpeg** - Audio format conversion - **Pandas** - Data processing ## 📊 Output Formats ### 1. Timestamped Transcript ``` AUDIO MINUTE | ROLE | TRANSCRIBED VERSION 00:00 - 01:00 min | ❓ INTERVIEWER | Q36. How many household members... 01:00 - 02:00 min | 💭 RESPONDENT | Three people ``` ### 2. English Translation ``` AUDIO MINUTE | ROLE | TRANSLATED VERSION 00:00 - 01:00 min | ❓ INTERVIEWER | Q36. How many household members... 01:00 - 02:00 min | 💭 RESPONDENT | Three people ``` ### 3. CSV Export All results can be downloaded as CSV files for easy import into Excel or databases. ## 🎯 Installation ### Prerequisites - Python 3.11 or higher - FFmpeg (for audio conversion) ### Setup 1. Clone the repository: ```bash git clon …