Update app.py

app.py

@@ -1,54 +1,76 @@
 import streamlit as st
-import pyttsx3  # For text-to-speech
+import pyttsx3  # For text-to-speech (offline)
 import speech_recognition as sr  # For speech recognition
-from gtts import gTTS  # Google Text-to-Speech for TTS (Alternative)
+from gtts import gTTS  # Google Text-to-Speech for TTS (alternative)
+import sounddevice as sd  # For audio recording
+import numpy as np  # Required for sounddevice
 import tempfile
 import os
 
 # Initialize text-to-speech engine (pyttsx3 for offline use)
 engine = pyttsx3.init()
 
-# Function to generate speech from text
-def speak_text(text):
-    # Using gTTS for an easy-to-use Google Text-to-Speech implementation
-    tts = gTTS(text=text, lang='en')
-    tts.save("sentence.mp3")
-    # Play the sentence using the default media player
-    os.system("start sentence.mp3")  # For Windows; use 'afplay sentence.mp3' for macOS, 'mpg321 sentence.mp3' for Linux
+# Function to generate speech from text (using pyttsx3 or gTTS)
+def speak_text(text, use_gtts=False):
+    if use_gtts:
+        # Use Google TTS (Online)
+        tts = gTTS(text=text, lang='en')
+        # Save to temporary file
+        with tempfile.NamedTemporaryFile(delete=False, suffix='.mp3') as tmpfile:
+            tts.save(tmpfile.name)
+            os.system(f"start {tmpfile.name}")  # For Windows; use 'afplay' for macOS, 'mpg321' for Linux
+    else:
+        # Use pyttsx3 (Offline)
+        engine.say(text)
+        engine.runAndWait()
+
+# Function to record user's speech using sounddevice
+def record_speech(duration=5, fs=16000):
+    # Record audio for a given duration (seconds) and sampling rate (fs)
+    st.write("Recording...")
+    audio_data = sd.rec(int(duration * fs), samplerate=fs, channels=1, dtype='int16')
+    sd.wait()  # Wait until recording is finished
+    return audio_data
 
-# Function to record user's speech
-def record_speech():
+# Function to transcribe speech using SpeechRecognition
+def transcribe_speech(audio_data, fs=16000):
     recognizer = sr.Recognizer()
-    with sr.Microphone() as source:
-        st.write("Listening...")
-        audio = recognizer.listen(source)
-        try:
-            st.write("Recognizing speech...")
-            text = recognizer.recognize_google(audio)
-            return text
-        except sr.UnknownValueError:
-            return "Sorry, I couldn't understand the speech."
-        except sr.RequestError:
-            return "Sorry, there was an error with the speech recognition service."
-
-# Streamlit UI
-st.title("English Shadowing Practice")
-
-# Input box for the user to type a sentence
-sentence = st.text_input("Enter a sentence to practice")
+    with tempfile.NamedTemporaryFile(delete=False, suffix='.wav') as tmpfile:
+        # Save recorded audio to temporary file
+        sd.write(tmpfile.name, audio_data, fs)
+        with sr.AudioFile(tmpfile.name) as source:
+            audio = recognizer.record(source)  # Read entire audio file
+    try:
+        # Perform speech recognition
+        return recognizer.recognize_google(audio)
+    except sr.UnknownValueError:
+        return "Sorry, I could not understand the audio."
+    except sr.RequestError as e:
+        return f"Could not request results from Google Speech Recognition service; {e}"
+
+# Streamlit App UI
+st.title("English Shadowing App")
+st.write("Practice speaking English by shadowing.")
+
+# Get user input sentence
+sentence = st.text_input("Enter a sentence to practice:")
 
 if sentence:
-    # Display the input sentence
-    st.write(f"Your sentence: {sentence}")
-    # Play the sentence
-    speak_text(sentence)
-
-# Button to record user's speech
-if st.button("Record Your Speech"):
-    speech = record_speech()
-    st.write(f"You said: {speech}")
-    # Provide feedback or allow comparison
-    if speech.lower() == sentence.lower():
-        st.write("Good job! Your speech matches the sentence.")
-    else:
-        st.write("Try again!")
+    # Play the sentence using TTS
+    speak_text(sentence, use_gtts=True)
+
+    # Record user's speech
+    st.write("Now, please say the sentence.")
+    if st.button("Start Recording"):
+        audio_data = record_speech(duration=5)
+        st.write("Recording complete.")
+
+        # Transcribe and compare
+        transcription = transcribe_speech(audio_data)
+        st.write(f"Your transcription: {transcription}")
+
+        # Show comparison
+        if transcription.lower() == sentence.lower():
+            st.success("Great job! Your pronunciation is correct.")
+        else:
+            st.warning("Try again. Keep practicing!")