Update app.py

app.py

@@ -2,6 +2,7 @@ from flask import Flask, render_template, request, jsonify
 import os
 import torch
 import re
+import ffmpeg  # Ensure FFmpeg is installed
 from transformers import pipeline
 from gtts import gTTS
 from pydub import AudioSegment
@@ -44,6 +45,14 @@ SYMBOL_MAPPING = {
     "space": " "
 }
 
+# Function to convert audio to WAV format (Fix FFmpeg error)
+def convert_to_wav(input_path, output_path):
+    try:
+        audio = AudioSegment.from_file(input_path)
+        audio.export(output_path, format="wav")
+    except Exception as e:
+        raise Exception(f"Audio conversion failed: {str(e)}")
+
 # Function to extract meaningful words (Removes unnecessary phrases)
 def extract_meaningful_text(text):
     text = text.lower().strip()
@@ -75,16 +84,20 @@ def transcribe():
         return jsonify({"error": "No audio file provided"}), 400
 
     audio_file = request.files["audio"]
-    audio_path = os.path.join("static", "temp.wav")
-    audio_file.save(audio_path)
+    input_audio_path = os.path.join("static", "temp_input")
+    output_audio_path = os.path.join("static", "temp.wav")
+    audio_file.save(input_audio_path)
 
     try:
+        # Convert audio to proper WAV format
+        convert_to_wav(input_audio_path, output_audio_path)
+
         # Check if the audio contains valid speech
-        if is_silent_audio(audio_path):
+        if is_silent_audio(output_audio_path):
             return jsonify({"error": "No speech detected. Please try again."}), 400
         
         # Transcribe using Whisper
-        result = asr_model.transcribe(audio_path, language="en")
+        result = asr_model.transcribe(output_audio_path, language="en")
         transcribed_text = clean_transcription(result["text"])
         
         return jsonify({"text": transcribed_text})
@@ -93,4 +106,4 @@ def transcribe():
 
 # Use Waitress for Production Server
 if __name__ == "__main__":
-    serve(app, host="0.0.0.0", port=7860)
+    serve(app, host="0.0.0.0", port=7860)