app.py

improve the text in the rcommendation, display each point on next line and improve the UI a little bit

import os
import requests
import pandas as pd
import numpy as np
import joblib
import gradio as gr
from datetime import datetime, timedelta
from tensorflow.keras.models import load_model
from tensorflow.keras.preprocessing import image as keras_image
from tensorflow.keras.applications.vgg16 import preprocess_input as vgg_preprocess
from tensorflow.keras.applications.xception import preprocess_input as xce_preprocess
from tensorflow.keras.losses import BinaryFocalCrossentropy
from PIL import Image

# --- CONFIGURATION ---
FOREST_COORDS = {'Pakistan Forest': (34.0, 73.0)}
API_URL = (
    "https://archive-api.open-meteo.com/v1/archive"
    "?latitude={lat}&longitude={lon}"
    "&start_date={start}&end_date={end}"
    "&daily=temperature_2m_max,temperature_2m_min,"
    "precipitation_sum,windspeed_10m_max,"
    "relative_humidity_2m_max,relative_humidity_2m_min"
    "&timezone=UTC"
)

# --- LOAD MODELS ---
def load_models():
    # Fire detector (VGG16)
    vgg_model = load_model(
        'vgg16_focal_unfreeze_more.keras',
        custom_objects={'BinaryFocalCrossentropy': BinaryFocalCrossentropy}
    )
    # Severity classifier (Xception)
    def focal_loss_fixed(gamma=2., alpha=.25):
        import tensorflow.keras.backend as K
        def loss_fn(y_true, y_pred):
            eps = K.epsilon(); y_pred = K.clip(y_pred, eps, 1.-eps)
            ce = -y_true * K.log(y_pred)
            w = alpha * K.pow(1-y_pred, gamma)
            return K.mean(w * ce, axis=-1)
        return loss_fn
    xce_model = load_model(
        'severity_post_tta.keras',
        custom_objects={'focal_loss_fixed': focal_loss_fixed()}
    )
    # Ensemble and trend models
    rf_model = joblib.load('ensemble_rf_model.pkl')
    xgb_model = joblib.load('ensemble_xgb_model.pkl')
    lr_model = joblib.load('wildfire_logistic_model_synthetic.joblib')
    return vgg_model, xce_model, rf_model, xgb_model, lr_model

vgg_model, xception_model, rf_model, xgb_model, lr_model = load_models()

# --- RULES & TEMPLATES ---
target_map = {0: 'mild', 1: 'moderate', 2: 'severe'}
trend_map = {1: 'increase', 0: 'same', -1: 'decrease'}
task_rules = {
    'mild':    {'decrease':'mild','same':'mild','increase':'moderate'},
    'moderate':{'decrease':'mild','same':'moderate','increase':'severe'},
    'severe':  {'decrease':'moderate','same':'severe','increase':'severe'}
}
templates = {
    'mild': (
        "📌 **Immediate Monitoring:** Although fire intensity is low, assign lookouts to monitor hotspots every 30 minutes. Use handheld IR cameras to detect any hidden flare-ups.\n\n"
        "📌 **Community Alert:** Send SMS alerts to nearby villages reminding them to stay vigilant. Provide clear instructions on how to report any smoke sightings.\n\n"
        "📌 **Fuel Management:** Conduct targeted removal of leaf litter and dry underbrush within a 100 m radius to reduce the chance of flare-ups.\n\n"
        "📌 **Preparedness Drills:** Hold a quick drill with ground crews to review communication protocols and ensure equipment (hoses, pumps) is ready.\n\n"
        "📌 **Public Education:** Distribute flyers on safe fire-watch practices and set up a hotline for rapid reporting."
    ),
    'moderate': (
        "🚒 **Rapid Response:** Dispatch two engine crews and one aerial water-drop helicopter. Coordinate with the regional command center to stage retardant tanks nearby.\n\n"
        "🏃‍♂️ **Evacuation Prep:** Pre-position evacuation buses at community centers. Issue voluntary evacuation notices to residents within 5 km downwind.\n\n"
        "🛠 **Containment Lines:** Construct a 10 m fire break using both hand tools and bulldozers. Apply fire-retardant gel along the anticipated flank.\n\n"
        "🌱 **Fuel Reduction:** Begin mechanical thinning of small trees and brush in high-risk zones adjacent to critical infrastructure.\n\n"
        "📣 **Awareness Campaign:** Launch radio spots explaining what to do if fire approaches, including evacuation routes and shelter locations."
    ),
    'severe': (
        "✈️ **Full Suppression:** Mobilize two air tankers for retardant drops and four ground crews with heavy equipment. Integrate real-time satellite imagery for targeting.\n\n"
        "🚨 **Mandatory Evacuation:** Issue immediate evacuation orders for all residents within a 10 km radius. Open three emergency shelters with medical staff on standby.\n\n"
        "🔥 **Backfire Operations:** Conduct controlled backfires under supervision of senior incident commanders to remove fuel ahead of the main front.\n\n"
        "🌳 **Post-Fire Rehabilitation:** Plan reforestation with fire-resistant native species; stabilize soil to prevent erosion in burn scar areas.\n\n"
        "🗣 **Crisis Communication:** Hold daily press briefings and social media updates. Provide mental-health support hotlines for displaced families."
    )
}

# --- PIPELINE FUNCTIONS ---
def detect_fire(img):
    x = keras_image.img_to_array(img.resize((128,128)))[None]
    x = vgg_preprocess(x)
    prob = float(vgg_model.predict(x)[0][0])
    return prob >= 0.5, prob


def classify_severity(img):
    x = keras_image.img_to_array(img.resize((224,224)))[None]
    x = xce_preprocess(x)
    preds = xception_model.predict(x)
    rf_p = rf_model.predict(preds)[0]
    xgb_p = xgb_model.predict(preds)[0]
    ensemble = int(round((rf_p + xgb_p)/2))
    return target_map.get(ensemble, 'moderate')


def fetch_weather_trend(lat, lon):
    end = datetime.utcnow()
    start = end - timedelta(days=1)
    url = API_URL.format(lat=lat, lon=lon,
                         start=start.strftime('%Y-%m-%d'),
                         end=end.strftime('%Y-%m-%d'))
    df = pd.DataFrame(requests.get(url).json().get('daily', {}))
    for c in ['precipitation_sum','temperature_2m_max','temperature_2m_min',
              'relative_humidity_2m_max','relative_humidity_2m_min','windspeed_10m_max']:
        df[c] = pd.to_numeric(df.get(c,[]), errors='coerce')
    df['precipitation'] = df['precipitation_sum'].fillna(0)
    df['temperature'] = (df['temperature_2m_max'] + df['temperature_2m_min'])/2
    df['humidity']    = (df['relative_humidity_2m_max'] + df['relative_humidity_2m_min'])/2
    df['wind_speed']  = df['windspeed_10m_max']
    df['fire_risk_score'] = (
        0.4*(df['temperature']/55) +
        0.2*(1-df['humidity']/100) +
        0.3*(df['wind_speed']/60) +
        0.1*(1-df['precipitation']/50)
    )
    feats = df[['temperature','humidity','wind_speed','precipitation','fire_risk_score']]
    feat = feats.fillna(feats.mean()).iloc[-1].values.reshape(1,-1)
    trend_cl = lr_model.predict(feat)[0]
    return trend_map.get(trend_cl, 'same')


def generate_recommendations(original_severity, weather_trend):
    # determine projected severity
    proj = task_rules[original_severity][weather_trend]
    rec = templates[proj]
    # proper multi-line header
    header = f"""**Original:** {original_severity.title()}  
**Trend:** {weather_trend.title()}  
**Projected:** {proj.title()}\n\n"""
    return header + rec

# --- GRADIO INTERFACE ---
def pipeline(image):
    img = Image.fromarray(image).convert('RGB')
    fire, prob = detect_fire(img)
    if not fire:
        return f"No wildfire detected (prob={prob:.2f})", "N/A", "N/A", "**No wildfire detected. Stay alert.**"
    sev = classify_severity(img)
    trend = fetch_weather_trend(*FOREST_COORDS['Pakistan Forest'])
    recs = generate_recommendations(sev, trend)
    return f"Fire Detected (prob={prob:.2f})", sev.title(), trend, recs

interface = gr.Interface(
    fn=pipeline,
    inputs=gr.Image(type='numpy', label='Upload Wildfire Image'),
    outputs=[
        gr.Textbox(label='Fire Status'),
        gr.Textbox(label='Severity Level'),
        gr.Textbox(label='Weather Trend'),
        gr.Markdown(label='Recommendations')
    ],
    title='Wildfire Detection & Management Assistant',
    description='Upload an image from a forest region in Pakistan to determine wildfire presence, severity, weather-driven trend, projection, and get expert recommendations.'
)

if __name__ == '__main__':
    interface.launch()