Update app.py

app.py

@@ -2,6 +2,7 @@ import gradio as gr
 import torch
 import logging
 import numpy as np
+import os
 from PIL import Image
 from transformers import ViTForImageClassification, ViTImageProcessor
 
@@ -23,15 +24,26 @@ except Exception as e:
     logger.error(f"Failed to load model: {str(e)}")
     raise RuntimeError(f"Model initialization failed: {str(e)}")
 
+def get_filename(image_path):
+    """Helper function to safely get a filename regardless of input type"""
+    if hasattr(image_path, 'name'):
+        return image_path.name
+    elif isinstance(image_path, str):
+        return os.path.basename(image_path)
+    else:
+        return "unknown_image"
+
 def preprocess_image(image_path):
     """Preprocess image for model input with proper error handling"""
     try:
+        # Handle both string paths and file objects
         pil_image = Image.open(image_path).convert("RGB")
         # Resize while maintaining aspect ratio
         width, height = pil_image.size
         new_size = (224, 224)
         pil_image = pil_image.resize(new_size, Image.Resampling.LANCZOS)
-        logger.info(f"Successfully preprocessed image: {image_path.name} ({width}x{height} → 224x224)")
+        filename = get_filename(image_path)
+        logger.info(f"Successfully preprocessed image: {filename} ({width}x{height} → 224x224)")
         return pil_image
     except Exception as e:
         logger.error(f"Image preprocessing error: {str(e)}")
@@ -93,7 +105,8 @@ def detect(image, confidence_threshold=0.7, detailed_analysis=False):
             feature_analysis = analyze_facial_features(pil_image, probabilities)
         
         # Logging for diagnostics and auditing
-        logger.info(f"Analysis results for {image.name}:")
+        filename = get_filename(image)
+        logger.info(f"Analysis results for {filename}:")
         logger.info(f"  - Raw probabilities: Real={confidence_real:.2f}%, Fake={confidence_fake:.2f}%")
         logger.info(f"  - Threshold ({confidence_threshold}): Predicted as {threshold_predicted}")