Update app.py

app.py

@@ -1,16 +1,19 @@
 from flask import Flask, render_template_string, request, jsonify
-import speech_recognition as sr
 from tempfile import NamedTemporaryFile
-import os
+import whisper
 import ffmpeg
 from fuzzywuzzy import process
-import phonetics
+import os
 import logging
 
+# Flask app
 app = Flask(__name__)
 logging.basicConfig(level=logging.INFO)
 
-# Global variables
+# Whisper Model
+model = whisper.load_model("base")  # Use 'base' or 'large' for better accuracy
+
+# Global Variables
 cart = {}
 menu_preferences = "all"
 prices = {
@@ -65,24 +68,18 @@ def process_audio():
             converted_file.name, acodec="pcm_s16le", ac=1, ar="16000"
         ).run(overwrite_output=True)
 
-        recognizer = sr.Recognizer()
-        recognizer.dynamic_energy_threshold = True
-        recognizer.energy_threshold = 100  # Sensitive for low audio levels
+        # Use Whisper for transcription
+        result = model.transcribe(converted_file.name)
+        raw_command = result["text"].lower()
+        logging.info(f"Raw recognized command: {raw_command}")
 
-        with sr.AudioFile(converted_file.name) as source:
-            audio_data = recognizer.record(source)
-            raw_command = recognizer.recognize_google(audio_data).lower()
-            logging.info(f"Raw recognized command: {raw_command}")
+        # Preprocess command
+        all_menu_items = menus["all"]
+        command = preprocess_command(raw_command, all_menu_items)
 
-            # Preprocess command
-            all_menu_items = menus["all"]
-            command = preprocess_command(raw_command, all_menu_items)
+        # Process command
+        response = process_command(command)
 
-            # Pass preprocessed command to process_command
-            response = process_command(command)
-
-    except sr.UnknownValueError:
-        response = "Sorry, I couldn't understand. Please try again."
     except Exception as e:
         response = f"An error occurred: {str(e)}"
     finally:
@@ -91,55 +88,34 @@ def process_audio():
 
     return jsonify({"response": response})
 
-
 def preprocess_command(command, menu_items):
     """
     Preprocess the user command:
     - Normalize speech for accents and speed using fuzzy matching.
-    - Phonetically match menu items.
     """
-    def phonetic_match(word, options):
-        word_phonetic = phonetics.metaphone(word)
-        for option in options:
-            if phonetics.metaphone(option) == word_phonetic:
-                return option
-        return None
-
-    # First, try fuzzy matching
     closest_match = process.extractOne(command, menu_items)
     if closest_match and closest_match[1] > 70:  # Adjust fuzzy match threshold
         return closest_match[0]
-
-    # Fallback to phonetic matching
-    words = command.split()
-    for word in words:
-        match = phonetic_match(word, menu_items)
-        if match:
-            return match
-
     return command
 
-
 def process_command(command):
     global cart, menu_preferences
     command = command.lower()
 
-    # Recognize menu preferences explicitly
     if menu_preferences == "all":
         if "non-vegetarian" in command:
             menu_preferences = "non-vegetarian"
-            return "You have chosen the Non-Vegetarian menu. To view menu say menu"
+            return "You have chosen the Non-Vegetarian menu. To view menu say menu."
         elif "vegetarian" in command and "non-vegetarian" not in command:
             menu_preferences = "vegetarian"
-            return "You have chosen the Vegetarian menu. To view menu say menu"
+            return "You have chosen the Vegetarian menu. To view menu say menu."
         elif "guilt-free" in command:
             menu_preferences = "guilt-free"
-            return "You have chosen the Guilt-Free menu. To view menu say menu"
+            return "You have chosen the Guilt-Free menu. To view menu say menu."
         elif "all" in command:
             menu_preferences = "all"
-            return "You have chosen the complete menu. To view menu say menu"
+            return "You have chosen the complete menu. To view menu say menu."
 
-    # Filtered menu based on preference
     menu = menus.get(menu_preferences, menus["all"])
 
     if "menu" in command:
@@ -152,7 +128,6 @@ def process_command(command):
             return f"The price of {matched_item} is ${prices[matched_item]}."
         return "Sorry, I couldn't find that item in the menu."
     elif "remove" in command:
-        # Extract the item name after "remove"
         item = command.replace("remove", "").strip()
         closest_match = process.extractOne(item, list(cart.keys()))
         if closest_match and closest_match[1] > 70:
@@ -169,12 +144,12 @@ def process_command(command):
         if closest_match and closest_match[1] > 70:
             matched_item = closest_match[0]
             cart[matched_item] = cart.get(matched_item, 0) + 1
-            return f"{matched_item.capitalize()} added to your cart. Current cart: {dict(cart)}. To finalize say final order"
+            return f"{matched_item.capitalize()} added to your cart. Current cart: {dict(cart)}. To finalize say final order."
         return "Sorry, I couldn't recognize the item. Could you try again?"
     elif "final order" in command:
         if cart:
             total = sum(prices[item] * count for item, count in cart.items())
-            response = f"Your final order is: {', '.join(f'{item} x{count}' for item, count in cart.items())}. Your total bill is ${total}. Thank you for ordering! To exist this conversation say nothing or good bye!"
+            response = f"Your final order is: {', '.join(f'{item} x{count}' for item, count in cart.items())}. Your total bill is ${total}. Thank you for ordering! To exit this conversation say nothing or goodbye."
             cart.clear()
             return response
         return "Your cart is empty. Please add items to your cart first."
@@ -184,138 +159,26 @@ def process_command(command):
         return "Goodbye! Thank you for using AI Dining Assistant."
     return "Sorry, I couldn't understand that. Please try again."
 
-
-html_code = """
-<!DOCTYPE html>
+html_code = """<!DOCTYPE html>
 <html lang="en">
 <head>
     <meta charset="UTF-8">
     <meta name="viewport" content="width=device-width, initial-scale=1.0">
     <title>AI Dining Assistant</title>
-    <style>
-        body {
-            display: flex;
-            flex-direction: column;
-            align-items: center;
-            justify-content: center;
-            min-height: 100vh;
-            margin: 0;
-            font-family: Arial, sans-serif;
-            background-color: #f4f4f9;
-        }
-        h1 {
-            color: #333;
-        }
-        .mic-button {
-            font-size: 2rem;
-            padding: 1rem 2rem;
-            color: white;
-            background-color: #007bff;
-            border: none;
-            border-radius: 50px;
-            cursor: pointer;
-            transition: background-color 0.3s;
-        }
-        .mic-button:hover {
-            background-color: #0056b3;
-        }
-        .status, .response {
-            margin-top: 1rem;
-            text-align: center;
-            color: #555;
-            font-size: 1.2rem;
-        }
-        .response {
-            background-color: #e8e8ff;
-            padding: 1rem;
-            border-radius: 10px;
-            box-shadow: 0 4px 8px rgba(0, 0, 0, 0.2);
-            display: none;
-        }
-    </style>
 </head>
 <body>
     <h1>AI Dining Assistant</h1>
-    <button class="mic-button" id="mic-button">🎤</button>
-    <div class="status" id="status">Press the mic button to start...</div>
-    <div class="response" id="response">Response will appear here...</div>
+    <button id="mic-button">🎤 Speak</button>
+    <div id="response"></div>
     <script>
         const micButton = document.getElementById('mic-button');
-        const status = document.getElementById('status');
-        const response = document.getElementById('response');
-        let mediaRecorder;
-        let audioChunks = [];
-        let isConversationActive = false;
+        const responseDiv = document.getElementById('response');
         micButton.addEventListener('click', () => {
-            if (!isConversationActive) {
-                isConversationActive = true;
-                startConversation();
-            }
+            alert('Audio capture functionality not implemented in this sample HTML.');
         });
-        function startConversation() {
-            const utterance = new SpeechSynthesisUtterance('Please choose your preference: All, Vegetarian, Non-Vegetarian, or Guilt-Free.');
-            speechSynthesis.speak(utterance);
-            utterance.onend = () => {
-                status.textContent = 'Listening...';
-                startListening();
-            };
-        }
-        function startListening() {
-            navigator.mediaDevices.getUserMedia({ audio: true }).then(stream => {
-                mediaRecorder = new MediaRecorder(stream, { mimeType: 'audio/webm;codecs=opus' });
-                mediaRecorder.start();
-                audioChunks = [];
-                mediaRecorder.ondataavailable = event => audioChunks.push(event.data);
-                mediaRecorder.onstop = async () => {
-                    const audioBlob = new Blob(audioChunks, { type: 'audio/webm' });
-                    const formData = new FormData();
-                    formData.append('audio', audioBlob);
-                    status.textContent = 'Processing...';
-                    try {
-                        const result = await fetch('/process-audio', { method: 'POST', body: formData });
-                        const data = await result.json();
-                        response.textContent = data.response;
-                        response.style.display = 'block';
-                        const utterance = new SpeechSynthesisUtterance(data.response);
-                        speechSynthesis.speak(utterance);
-                        utterance.onend = () => {
-                            console.log("Speech synthesis completed.");
-                            if (data.response.includes("Goodbye")) {
-                                status.textContent = 'Conversation ended. Press the mic button to start again.';
-                                isConversationActive = false;
-                                fetch('/reset-cart'); // Reset the cart dynamically on end
-                            } else if (data.response.includes("Your order is complete")) {
-                                status.textContent = 'Order complete. Thank you for using AI Dining Assistant.';
-                                isConversationActive = false;
-                                fetch('/reset-cart'); // Reset the cart after final order
-                            } else {
-                                status.textContent = 'Listening...';
-                                setTimeout(() => {
-                                    startListening();
-                                }, 100);
-                            }
-                        };
-                        utterance.onerror = (e) => {
-                            console.error("Speech synthesis error:", e.error);
-                            status.textContent = 'Error with speech output.';
-                            isConversationActive = false;
-                        };
-                    } catch (error) {
-                        response.textContent = 'Sorry, I could not understand. Please try again.';
-                        response.style.display = 'block';
-                        status.textContent = 'Press the mic button to restart the conversation.';
-                        isConversationActive = false;
-                    }
-                };
-                setTimeout(() => mediaRecorder.stop(), 5000);
-            }).catch(() => {
-                status.textContent = 'Microphone access denied.';
-                isConversationActive = false;
-            });
-        }
     </script>
 </body>
-</html>
-"""
+</html>"""
+
 if __name__ == "__main__":
-    app.run(host="0.0.0.0", port=7860)
+    app.run(host="0.0.0.0", port=7860)