Upload 3 files

app.py

@@ -0,0 +1,114 @@
+# -*- coding: utf-8 -*-
+"""Untitled3.ipynb
+
+Automatically generated by Colab.
+
+Original file is located at
+    https://colab.research.google.com/drive/15hJqN0ojQZbWh5t6vlnLXzMAAuUyY84Y
+"""
+
+import gradio as gr
+import pandas as pd
+import tensorflow as tf
+import numpy as np
+import requests
+from io import BytesIO
+from tensorflow.keras.applications.mobilenet_v2 import MobileNetV2, preprocess_input
+from tensorflow.keras.preprocessing import image
+
+# Load model and data
+model = tf.keras.models.load_model("model.h5")
+df = pd.read_csv("skincare_products_clean.csv")
+
+# Load MobileNetV2 feature extractor
+feature_extractor = MobileNetV2(weights='imagenet', include_top=False, pooling='avg')
+
+# Reduction layer to match model input (32 features)
+reduce_model = tf.keras.Sequential([
+    tf.keras.layers.InputLayer(input_shape=(1280,)),
+    tf.keras.layers.Dense(32, activation='relu')
+])
+
+CLASS_NAMES = [
+    "acne", "aging", "dryness", "sensitivity", "pigmentation",
+    "eczema", "rosacea", "dark circles", "wrinkles", "scars"
+]
+
+concern_ingredients = {
+    "acne": ["salicylic acid", "benzoyl peroxide", "niacinamide", "azelaic acid", "tea tree"],
+    "dryness": ["hyaluronic acid", "glycerin", "ceramide", "squalane"],
+    "aging": ["retinol", "peptides", "vitamin c", "niacinamide"],
+    "sensitivity": ["allantoin", "panthenol", "madecassoside", "centella"],
+    "pigmentation": ["vitamin c", "kojic acid", "azelaic acid", "niacinamide"],
+}
+
+def recommend_products(concern, num_results=5, max_price=None, sort_by="price"):
+    ingredients = concern_ingredients.get(concern.lower(), [concern.lower()])
+    pattern = '|'.join(ingredients)
+    df_filtered = df[df['clean_ingreds'].str.contains(pattern, case=False, na=False)]
+
+    if max_price:
+        df_filtered = df_filtered[df_filtered['price'] <= float(max_price)]
+
+    if df_filtered.empty:
+        df_filtered = df.copy()
+
+    if sort_by in df_filtered.columns:
+        df_filtered = df_filtered.sort_values(by=sort_by)
+    else:
+        df_filtered = df_filtered.sort_values(by="price")
+
+    df_filtered['clean_ingreds'] = df_filtered['clean_ingreds'].apply(
+        lambda x: ', '.join(eval(x)[:3]) + "..." if isinstance(x, str) and x.startswith('[') else x
+    )
+    df_filtered['price'] = df_filtered['price'].round(2)
+
+    output = ""
+    for _, row in df_filtered.head(num_results).iterrows():
+        output += f"🧴 **Product**: {row['product_name']}\n"
+        output += f"📦 **Type**: {row['product_type']}\n"
+        output += f"🧪 **Key Ingredients**: {row['clean_ingreds']}\n"
+        output += f"💰 **Price**: £{row['price']}\n\n"
+    return output.strip()
+
+def diagnose_image(image_input):
+    try:
+        img = image_input.resize((224, 224))
+        img_array = image.img_to_array(img)
+        img_array = np.expand_dims(img_array, axis=0)
+        img_array = preprocess_input(img_array)
+
+        features_1280 = feature_extractor.predict(img_array)
+        features_32 = reduce_model.predict(features_1280)
+        preds = model.predict(features_32)
+
+        class_index = np.argmax(preds[0])
+        confidence = preds[0][class_index] * 100
+        return f"✅ Predicted Diagnosis: **{CLASS_NAMES[class_index].capitalize()}**\n🔎 Confidence: {confidence:.2f}%"
+    except Exception as e:
+        return f"❌ Error: {e}"
+
+# Gradio Interface
+text_interface = gr.Interface(
+    fn=recommend_products,
+    inputs=[
+        gr.Textbox(label="Skincare Concern (e.g., acne, dryness)"),
+        gr.Number(label="Number of Recommendations", value=5),
+        gr.Number(label="Max Price (Optional)", value=None),
+        gr.Textbox(label="Sort By (price, product_name, etc.)", value="price")
+    ],
+    outputs="text",
+    title="🧴 Skincare Product Recommender"
+)
+
+image_interface = gr.Interface(
+    fn=diagnose_image,
+    inputs=gr.Image(type="pil", label="Upload Skin Image"),
+    outputs="text",
+    title="🧬 Skin Condition Diagnosis"
+)
+
+gr.TabbedInterface(
+    [text_interface, image_interface],
+    tab_names=["🧴 Text-Based Recommender", "🧬 Image-Based Diagnosis"]
+).launch()

model.h5

@@ -0,0 +1,3 @@
+version https://git-lfs.github.com/spec/v1
+oid sha256:039ef1e6f47274afa21df6cd06026b47fc79a288e2bd5e9ec7a168d90df875a7
+size 110320