Update app.py

app.py

@@ -7,6 +7,7 @@ import tempfile
 import shutil
 import re
 import speech_recognition as sr  # Library for voice recognition
+import sounddevice as sd
 
 def translate_to_japanese(api_key, text):
     """
@@ -77,16 +78,23 @@ def generate_audio_from_text(text):
     tts.save(temp_audio_file.name)
     return temp_audio_file.name
 
-# Function to capture and transcribe voice input
 def transcribe_voice_input():
     recognizer = sr.Recognizer()
-    with sr.Microphone() as source:
-        st.info("Please speak now...")
-        recognizer.adjust_for_ambient_noise(source)
-        audio = recognizer.listen(source)
+
+    # Set parameters for sounddevice (e.g., 44100 Hz, 1 channel)
+    duration = 5  # Seconds to record
+    fs = 44100  # Sampling frequency
+
+    st.info("Please speak now...")
+
+    # Record the audio using sounddevice
+    audio_data = sd.rec(int(duration * fs), samplerate=fs, channels=1, dtype='int16')
+    sd.wait()  # Wait until recording is finished
+
+    # Convert the numpy array to audio and recognize the speech
+    with sr.AudioData(audio_data.tobytes(), fs, 2) as source:
         try:
-            # Use Google Web Speech API to transcribe audio
-            text = recognizer.recognize_google(audio)
+            text = recognizer.recognize_google(source)
             st.success(f"Transcribed text: {text}")
             return text
         except sr.UnknownValueError: