app.py

# app.py
import streamlit as st
from PIL import Image
import tensorflow as tf
import numpy as np
from tensorflow.keras.applications.efficientnet import preprocess_input
import json
from io import BytesIO
import google.generativeai as genai
from dotenv import load_dotenv
import os

# Load environment variables
load_dotenv()
GOOGLE_API_KEY = os.getenv("GOOGLE_API_KEY")
genai.configure(api_key=GOOGLE_API_KEY)

# Load model
IMG_SIZE = (299, 299)
model = tf.keras.models.load_model("model.h5")
with open("trained_classes.json", "r") as f:
    class_labels = json.load(f)

# Streamlit UI
st.title("NutriSync: Food Nutrition Analyzer")
uploaded_file = st.file_uploader("Upload food image", type=["jpg","jpeg","png"])

def predict_food(img):
    img = img.resize(IMG_SIZE)
    img_array = np.array(img)
    if len(img_array.shape) == 2:
        img_array = np.stack((img_array,) * 3, axis=-1)
    if img_array.shape[-1] == 4:
        img_array = img_array[..., :3]
    img_array = preprocess_input(img_array)
    img_array = np.expand_dims(img_array, axis=0)
    preds = model.predict(img_array)
    idx = np.argmax(preds[0])
    confidence = float(preds[0][idx])
    food = class_labels[idx] if idx < len(class_labels) else "unknown"
    return food, confidence

def get_gemini_response(food_item, img_bytes, prompt):
    try:
        model_gemini = genai.GenerativeModel("gemini-1.5-flash-latest")
        response = model_gemini.generate_content([food_item, {"mime_type":"image/jpeg","data":img_bytes}, prompt])
        return response.text
    except Exception as e:
        return f"Error retrieving nutritional info: {e}"

if uploaded_file is not None:
    img = Image.open(uploaded_file).convert("RGB")
    st.image(img, caption="Uploaded Image", use_column_width=True)

    food, confidence = predict_food(img)
    st.write(f"Predicted Food: {food}, Confidence: {confidence:.2f}")

    # Reset file pointer for Gemini
    uploaded_file.seek(0)
    img_bytes = uploaded_file.read()

    prompt = (
        f"Analyze the nutritional content of {food}. "
        "Provide estimated values for Calories, Protein (g), Carbs (g), Fats (g), Fiber (g), Sugars (g) "
        "with short dietary recommendations."
        if food != "unknown" and confidence >= 0.6
        else "Analyze the nutritional content of food. Provide estimated values only."
    )

    nutrition_info = get_gemini_response(food, img_bytes, prompt)
    st.write("Nutritional Info:")
    st.write(nutrition_info)