Update app.py

app.py

@@ -2,9 +2,13 @@ import streamlit as st
 from transformers import pipeline
 import google.generativeai as genai
 from pytube import Search
+import speech_recognition as sr
+import tempfile
+from pydub import AudioSegment
+import numpy as np
 
 # Load sentiment analysis model using PyTorch backend
-mood_classifier = pipeline("sentiment-analysis", framework="pt")  # Use PyTorch
+mood_classifier = pipeline("sentiment-analysis", framework="pt")
 
 # Functions
 def detect_mood(text):
@@ -16,11 +20,46 @@ def detect_mood(text):
     else:
         return "neutral"
 
+def speech_to_text():
+    # Initialize recognizer
+    r = sr.Recognizer()
+    
+    # Create a temporary file to store the recorded audio
+    with tempfile.NamedTemporaryFile(delete=False, suffix=".wav") as fp:
+        # Add audio recording widget
+        audio_bytes = st.audio_recorder(
+            text="Click to record your mood",
+            recording_color="#e8b62c",
+            neutral_color="#6aa36f"
+        )
+        
+        if audio_bytes:
+            # Save audio bytes to temporary file
+            fp.write(audio_bytes)
+            temp_filename = fp.name
+            
+            # Read the audio file
+            with sr.AudioFile(temp_filename) as source:
+                # Adjust for ambient noise and record
+                r.adjust_for_ambient_noise(source)
+                audio = r.record(source)
+                
+                try:
+                    # Use Google Speech Recognition
+                    text = r.recognize_google(audio)
+                    return text
+                except sr.UnknownValueError:
+                    st.error("Could not understand the audio")
+                    return None
+                except sr.RequestError:
+                    st.error("Could not request results from speech recognition service")
+                    return None
+    return None
+
 def get_song_recommendations(mood, api_key):
     try:
         genai.configure(api_key=api_key)
         model = genai.GenerativeModel('gemini-pro')
-
         # System prompt to guide the AI
         system_prompt = """
         You are a music recommendation assistant specialized in Indian songs. Your task is to suggest popular Indian songs based on the user's mood.
@@ -59,16 +98,30 @@ mood_options = ["happy", "sad", "energetic", "romantic", "calm"]
 # Input for Gemini API key
 gemini_api_key = st.sidebar.text_input("Enter your Gemini API Key:", type="password")
 
-# Add a button for mood input
-user_mood = st.sidebar.selectbox("Select your mood:", mood_options)
+# Add option to choose between text and speech input
+input_method = st.sidebar.radio("Choose input method:", ["Text", "Speech"])
+
+if input_method == "Text":
+    # Text input
+    user_mood = st.sidebar.selectbox("Select your mood:", mood_options)
+else:
+    # Speech input
+    st.write("📢 Tell me about your day...")
+    spoken_text = speech_to_text()
+    
+    if spoken_text:
+        st.write(f"You said: {spoken_text}")
+        user_mood = detect_mood(spoken_text)
+    else:
+        user_mood = None
 
 # Playlist
 if 'playlist' not in st.session_state:
     st.session_state.playlist = []
 
 # Main content
-if user_mood and gemini_api_key:
-    mood = detect_mood(user_mood)
+if (user_mood or spoken_text) and gemini_api_key:
+    mood = detect_mood(user_mood if user_mood else spoken_text)
     st.write(f"🎭 Detected Mood: **{mood}**")
     
     st.write("🎵 Recommended Songs:")