from flask import Flask, render_template_string, request, jsonify
import speech_recognition as sr
from tempfile import NamedTemporaryFile
import os
import ffmpeg
import logging
from werkzeug.exceptions import BadRequest
# Initialize Flask App
app = Flask(__name__)
# Set up logging
logging.basicConfig(level=logging.INFO)
# Initialize conversation state
user_order = [] # Stores the current order
user_preferences = {} # Stores the customer's preferences
# HTML Template for Frontend
html_code = """
AI Dining Assistant
AI Dining Assistant
Press the mic button to start the conversation...
Response will appear here...
"""
@app.route('/')
def index():
return render_template_string(html_code)
@app.route('/process-audio', methods=['POST'])
def process_audio():
global user_order, user_preferences
try:
# Validate audio file
audio_file = request.files.get('audio')
if not audio_file:
raise BadRequest("No audio file provided.")
temp_file = NamedTemporaryFile(delete=False, suffix=".webm")
audio_file.save(temp_file.name)
logging.info(f"Saved input audio to {temp_file.name}")
if os.path.getsize(temp_file.name) == 0:
raise BadRequest("Uploaded audio file is empty.")
# Convert audio to PCM WAV format (16kHz, mono)
converted_file = NamedTemporaryFile(delete=False, suffix=".wav")
try:
ffmpeg.input(temp_file.name).output(
converted_file.name, acodec='pcm_s16le', ac=1, ar='16000'
).run(overwrite_output=True)
except Exception as ffmpeg_error:
logging.error(f"FFmpeg conversion error: {str(ffmpeg_error)}")
return jsonify({"response": "Audio conversion failed. Please try again."})
logging.info(f"Converted audio saved to {converted_file.name}")
# Recognize speech
recognizer = sr.Recognizer()
with sr.AudioFile(converted_file.name) as source:
audio_data = recognizer.record(source)
try:
command = recognizer.recognize_google(audio_data)
logging.info(f"Recognized command: {command}")
response = process_command(command)
except sr.UnknownValueError:
logging.error("Google Speech Recognition could not understand the audio")
response = "Sorry, I couldn't understand your request. Please try again."
except sr.RequestError as e:
logging.error(f"Error with Google Speech Recognition service: {e}")
response = "Sorry, there was an issue with the speech recognition service."
return jsonify({"response": response})
except BadRequest as br:
logging.error(f"Bad request error: {br}")
return jsonify({"response": f"Bad Request: {str(br)}"})
except Exception as e:
logging.error(f"Error processing audio: {e}")
return jsonify({"response": f"An error occurred: {str(e)}"})
finally:
# Clean up temporary files
try:
if os.path.exists(temp_file.name):
os.unlink(temp_file.name)
if os.path.exists(converted_file.name):
os.unlink(converted_file.name)
except Exception as cleanup_error:
logging.error(f"Error cleaning up files: {cleanup_error}")
def process_command(command):
global user_order, user_preferences
command = command.lower()
if "hello" in command or "hi" in command or "hey" in command:
return (
"Welcome! How can I assist you with your meal today? "
"Please let me know your preferences."
)
elif "reset preferences" in command:
user_order = [] # Reset the order
user_preferences = {} # Reset preferences
return "Your preferences have been reset. What would you like to order?"
elif "show my order" in command or "what's my order" in command:
if user_order:
return "Your current order includes: " + ", ".join(user_order)
else:
return "You haven't added anything to your order yet."
elif "place order" in command or "confirm order" in command:
if user_order:
return (
"You have the following items in your order: " + ", ".join(user_order) +
". Would you like to confirm your order?"
)
else:
return "You haven't added anything to your order yet. Please add some items."
elif "yes" in command or "place order" in command:
return "Your order has been confirmed and sent to the kitchen. Thank you for ordering!"
return "Sorry, I didn't understand your request. You can ask to view your order, reset preferences, or place an order."
if __name__ == "__main__":
app.run(host="0.0.0.0", port=7860)