updated with voice the app.py

app.py

@@ -9,6 +9,9 @@ from transformers import pipeline, AutoTokenizer, AutoModelForSequenceClassifica
 from wordcloud import WordCloud
 import matplotlib.pyplot as plt
 import io
+import speech_recognition as sr
+from gtts import gTTS
+import os
 
 # Download NLTK resources
 nltk.download('punkt')
@@ -30,6 +33,22 @@ def load_classification_model():
 def load_qa_model():
     return pipeline("question-answering", model="deepset/roberta-base-squad2")
 
+def recognize_speech():
+    recognizer = sr.Recognizer()
+    with sr.Microphone() as source:
+        st.info("Listening... Speak now.")
+        try:
+            audio = recognizer.listen(source, timeout=5)  # Listen for 5 seconds
+            question_text = recognizer.recognize_google(audio)  # Convert speech to text
+            st.success(f"You said: {question_text}")  # Show recognized text
+            return question_text
+        except sr.UnknownValueError:
+            st.error("Sorry, could not understand the audio.")
+        except sr.RequestError:
+            st.error("Could not request results, check your internet connection.")
+    return None
+
+
 # Preprocessing function (same as in Section 01)
 def preprocess_text(text):
     # Lowercase
@@ -163,6 +182,10 @@ with tab2:
    
     
     question = st.text_input("Enter your question:")
+
+    if st.button("🎤 Speak"):
+        question = recognize_speech()
+        use_voice = True
     
     if st.button("Get Answer") and context and question:
         with st.spinner("Searching for answers..."):
@@ -174,6 +197,18 @@ with tab2:
             
             st.subheader("Details")
             st.write(f"Confidence: {result['score']:.2f}")
+
+
+    if use_voice:
+            tts = gTTS(result['answer'])  # Convert text answer to speech
+            temp_audio = tempfile.NamedTemporaryFile(delete=False, suffix=".mp3")  # Create temp file
+            tts.save(temp_audio.name)
+
+            # Play the Answer
+            st.audio(temp_audio.name, format="audio/mp3")  
+
+            # Cleanup temp file
+            os.remove(temp_audio.name)        
             
 
 with tab3: