Update src/streamlit_app.py

src/streamlit_app.py

@@ -2,6 +2,33 @@ import streamlit as st
 import whisper
 import tempfile
 
+import subprocess
+import tempfile
+import os
+
+st.logo(
+    LOGO_URL_LARGE,
+    link="https://jadavpuruniversity.in/storage/2021/12/logo-black-o.png",
+    icon_image=LOGO_URL_SMALL,
+)
+
+def convert_to_wav(input_path):
+    tmp_wav = tempfile.NamedTemporaryFile(delete=False, suffix=".wav")
+    tmp_wav.close()
+
+    cmd = [
+        "ffmpeg",
+        "-y",
+        "-i", input_path,
+        "-ar", "16000",   # sample rate
+        "-ac", "1",       # mono
+        tmp_wav.name
+    ]
+
+    subprocess.run(cmd, stdout=subprocess.DEVNULL, stderr=subprocess.DEVNULL, check=True)
+    return tmp_wav.name
+
+
 st.set_page_config(
     page_title="IndicASR",
     page_icon="✔️",
@@ -31,8 +58,10 @@ if choice == 'English':
         # Load Whisper model
         model = whisper.load_model("turbo")
 
+        wav_path = convert_to_wav(audio_path)
+
         # Transcribe
-        result = model.transcribe(audio_path, verbose=False)
+        result = model.transcribe(wav_path, verbose=False)
 
         st.subheader("Transcript")
         st.write(result["text"])