Updated app.py

app.py

@@ -1,28 +1,38 @@
-# app.py
-
-import streamlit as st
-from transformers import pipeline
-from PIL import Image
 import requests
+from io import BytesIO
+from PIL import Image
+import os
 
-# Load the Hugging Face pipeline for image classification
-classifier = pipeline("image-classification", model="imfarzanansari/skintelligent-acne")
-
-# Title of the Streamlit app
-st.title("Skin Condition Classification - Acne Detection")
+def get_acne_classification(image_input):
+    # Hugging Face Spaces URL
+    url = "https://huggingface.co/spaces/francoisMav/skin_acne"  # Update with your actual Space URL
 
-# Upload an image file
-uploaded_file = st.file_uploader("Choose an image...", type="jpg")
+    # Check if the input is a URL or a local file
+    if image_input.startswith('http://') or image_input.startswith('https://'):
+        # If it's a URL, download the image
+        response = requests.get(image_input)
+        if response.status_code == 200:
+            # Load the image into memory using PIL
+            image = Image.open(BytesIO(response.content))
+        else:
+            return {"error": "Failed to download image from URL"}
+    elif os.path.isfile(image_input):
+        # If it's a file path, open the image
+        image = Image.open(image_input)
+    else:
+        return {"error": "Invalid image input. Must be a URL or valid file path."}
 
-if uploaded_file is not None:
-    # Display the uploaded image
-    image = Image.open(uploaded_file)
-    st.image(image, caption="Uploaded Image", use_column_width=True)
+    # Convert image to the format needed for Hugging Face Spaces
+    buffered = BytesIO()
+    image.save(buffered, format="JPEG")
+    buffered.seek(0)
 
-    # Perform image classification using the Hugging Face pipeline
-    st.write("Classifying...")
-    predictions = classifier(image)
+    # Send the image to the Hugging Face Space via POST request
+    files = {'file': buffered}
+    response = requests.post(url, files=files)
 
-    # Display the results
-    for pred in predictions:
-        st.write(f"Label: {pred['label']}, Score: {pred['score']:.4f}")
+    # Check the response and return the prediction
+    if response.status_code == 200:
+        return response.json()  # Adjust based on how your Space returns predictions
+    else:
+        return {"error": "Failed to get prediction from Hugging Face Space"}