app.py

import streamlit as st
from pyannote.audio import Pipeline
from transformers import pipeline
import whisper

# Title
st.title("Multi-Speaker Audio Analyzer")

# Upload Audio File
uploaded_file = st.file_uploader("Upload an audio file (MP3/WAV)", type=["mp3", "wav"])

# Process Button
if uploaded_file:
    st.audio(uploaded_file, format='audio/wav')

    # Load pre-trained models
    diarization_pipeline = Pipeline.from_pretrained("pyannote/speaker-diarization")
    transcription_model = whisper.load_model("base")
    summarizer = pipeline("summarization", model="facebook/bart-large-cnn")

    # Perform Speaker Diarization
    st.write("Processing Speaker Diarization...")
    diarized_output = diarization_pipeline(uploaded_file)

    # Perform Speech-to-Text Transcription
    st.write("Transcribing Audio...")
    transcription = transcription_model.transcribe(uploaded_file)

    # Generate Summary
    st.write("Generating Summary...")
    summary = summarizer(transcription["text"])

    # Display Outputs
    st.write("Speaker-Diarized Transcript:")
    st.text(diarized_output)

    st.write("Full Transcription:")
    st.text(transcription["text"])

    st.write("Summary:")
    st.text(summary[0]['summary_text'])