Update app.py

app.py

@@ -3,95 +3,112 @@ from transformers import pipeline
 from PIL import Image
 import traceback
 
-# Correct model names
-MODELS = {
-    "umm-maybe": "umm-maybe/AI-image-detector",
-    "dima806": "dima806/ai_vs_human_generated_image_detection",
-    "ateeqq": "Ateeqq/ai-vs-human-image-detector"
-}
-
-# Equal weights
-WEIGHTS = {
-    "umm-maybe": 1,
-    "dima806": 1,
-    "ateeqq": 1
-}
-
-# Load pipelines
-pipes = {}
-for key, model_id in MODELS.items():
+# Define model priorities
+models = [
+    ("Ateeqq/ai-vs-human-image-detector", "ateeq"),
+    ("umm-maybe/AI-image-detector", "umm_maybe"),
+    ("dima806/ai_vs_human_generated_image_detection", "dimma"),
+]
+
+pipes = []
+for model_id, _ in models:
     try:
-        pipes[key] = pipeline("image-classification", model=model_id)
-        print(f"✅ Loaded {model_id}")
+        pipes.append((model_id, pipeline("image-classification", model=model_id)))
     except Exception as e:
-        print(f"⚠️ Failed to load {model_id}: {e}")
+        print(f"Error loading {model_id}: {e}")
 
 def predict_image(image: Image.Image):
     try:
-        scores = {}
         results = []
-
-        for key, pipe in pipes.items():
-            outputs = pipe(image)
-            ai_score = None
-
-            for o in outputs:
-                label = o['label'].lower()
-                score = o['score']
-
-                if 'ai' in label or 'artificial' in label or 'fake' in label:
-                    ai_score = score
-                    break
-                elif 'human' in label or 'real' in label:
-                    ai_score = 1 - score
-                    break
-
-            if ai_score is None:
-                ai_score = 0.5
-
-            scores[key] = ai_score
-            results.append({
-                "model": MODELS[key],
-                "ai_score": round(ai_score, 3)
-            })
-
-        # Weighted average
-        total_weight = sum(WEIGHTS.values())
-        final_score = sum(scores[k] * WEIGHTS[k] for k in scores) / total_weight
-        final_score = max(0, min(1, final_score))
-        final_percent = round(final_score * 100, 2)
-
-        verdict = "🧠 AI-Generated" if final_score > 0.55 else "🧍 Human"
-        color = "#ff4b4b" if final_score > 0.55 else "#4caf50"
-
-        # HTML output with bars
-        text = "## 🔍 Model Results\n"
-        for res in results:
-            percent = round(res["ai_score"] * 100, 2)
-            bar_color = "#ff4b4b" if percent > 50 else "#4caf50"
-            text += (
-                f"**{res['model']}** → {percent}% AI\n"
-                f"<div style='height:10px;background-color:{bar_color};width:{percent}%;border-radius:5px;margin-bottom:6px;'></div>\n"
-            )
-
-        text += f"\n## 🧾 Final Verdict\n"
-        text += f"<div style='font-size:18px;font-weight:bold;color:{color};'>{verdict} ({final_percent}%)</div>"
-
-        return text
+        for _, pipe in pipes:
+            res = pipe(image)[0]
+            results.append(res)
+
+        final_result = results[0]  # Highest priority model result
+        label = final_result["label"].lower()
+        score = final_result["score"] * 100
+
+        if "ai" in label or "fake" in label:
+            verdict = f"🧠 AI-Generated ({score:.1f}% confidence)"
+            color = "#007BFF"
+        else:
+            verdict = f"🧍 Human-Made ({score:.1f}% confidence)"
+            color = "#4CAF50"
+
+        html = f"""
+        <div class='result-box' style="
+            background: linear-gradient(135deg, {color}33, #1a1a1a);
+            border: 2px solid {color};
+            border-radius: 15px;
+            padding: 25px;
+            text-align: center;
+            color: white;
+            font-size: 20px;
+            font-weight: 600;
+            box-shadow: 0 0 20px {color}55;
+            animation: fadeIn 0.6s ease-in-out;
+        ">
+            {verdict}
+        </div>
+        """
+        return html
 
     except Exception as e:
         traceback.print_exc()
-        return f"❌ Error: {str(e)}"
-
-
-# Gradio app
-demo = gr.Interface(
-    fn=predict_image,
-    inputs=gr.Image(type="pil", label="Upload Image"),
-    outputs=gr.HTML(label="Results"),
-    title="🧩 Unreal Eye – AI Image Detector",
-    description="Detects AI-generated images using 3 equal-weight expert models (umm-maybe, dima806, ateeqq)."
-)
-
-if __name__ == "__main__":
-    demo.launch()
+        return f"<div style='color:red;'>Error analyzing image: {str(e)}</div>"
+
+# Custom CSS
+css = """
+.gr-button-primary {
+    background-color: #007BFF !important;
+    border-color: #007BFF !important;
+    color: white !important;
+    font-weight: 600;
+}
+.gr-button-secondary {
+    background-color: #dc3545 !important;
+    color: white !important;
+    font-weight: 600;
+}
+#loading-text {
+    text-align: center;
+    font-size: 18px;
+    color: #007BFF;
+    animation: blink 1.2s infinite;
+}
+@keyframes blink {
+    50% { opacity: 0.5; }
+}
+@keyframes fadeIn {
+    from { opacity: 0; transform: scale(0.95); }
+    to { opacity: 1; transform: scale(1); }
+}
+"""
+
+with gr.Blocks(css=css, theme=gr.themes.Soft()) as demo:
+    gr.Markdown("<h1 style='text-align:center; color:#007BFF;'>🔍 AI Image Detector</h1>")
+
+    with gr.Row():
+        with gr.Column(scale=1):
+            image_input = gr.Image(type="pil", label="Upload an image")
+            analyze_button = gr.Button("Analyze", variant="primary")
+            clear_button = gr.Button("Clear", variant="secondary")
+            loading_text = gr.Markdown("", elem_id="loading-text")
+        with gr.Column(scale=1):
+            output = gr.HTML(label="Result")
+
+    def analyze_with_animation(img):
+        loading_text.value = "⏳ Analyzing image..."
+        result = predict_image(img)
+        loading_text.value = ""
+        return result
+
+    analyze_button.click(
+        analyze_with_animation,
+        inputs=image_input,
+        outputs=output,
+    )
+
+    clear_button.click(lambda: (None, ""), outputs=[image_input, output])
+
+demo.launch()