Update app.py

Overwritting by local

app.py

@@ -33,6 +33,36 @@ model = GenerativeInferenceModel()
 
 # Define example images and their parameters with updated values from the research
 examples = [
+    {
+        "image": os.path.join("stimuli", "urbanoffice1.jpg"),
+        "name": "UrbanOffice1",
+        "wiki": "https://en.wikipedia.org/wiki/Visual_perception",
+        "papers": [
+            "[Adversarially Robust Vision](https://github.com/MadryLab/robustness)",
+            "[Generative Inference](https://doi.org/10.1016/j.tics.2003.08.003)"
+        ],
+        "method": "Prior-Guided Drift Diffusion",
+        "reverse_diff": {
+            "model": "resnet50_robust",
+            "layer": "all",
+            "initial_noise": 1.0,
+            "diffusion_noise": 0.002,
+            "step_size": 1.0,
+            "iterations": 500,
+            "epsilon": 40.0
+        },
+        "inference_normalization": "off",
+        "use_adaptive_eps": False,
+        "use_adaptive_step": True,
+        "mask_center_x": 0.5,
+        "mask_center_y": 0.0,
+        "mask_radius": 0.2,
+        "mask_sigma": 0.2,
+        "eps_max_mult": 20.0,
+        "eps_min_mult": 1.0,
+        "step_max_mult": 50.0,
+        "step_min_mult": 0.2,
+    },
     {
         "image": os.path.join("stimuli", "Neon_Color_Circle.jpg"),
         "name": "Neon Color Spreading",
@@ -185,6 +215,36 @@ examples = [
             "iterations": 101,
             "epsilon": 3.0
         }
+    },
+    {
+        "image": os.path.join("stimuli", "urbanoffice1.jpg"),
+        "name": "UrbanOffice1",
+        "wiki": "https://en.wikipedia.org/wiki/Visual_perception",
+        "papers": [
+            "[Adversarially Robust Vision](https://github.com/MadryLab/robustness)",
+            "[Generative Inference](https://doi.org/10.1016/j.tics.2003.08.003)"
+        ],
+        "method": "Prior-Guided Drift Diffusion",
+        "reverse_diff": {
+            "model": "resnet50_robust",
+            "layer": "all",
+            "initial_noise": 1.0,
+            "diffusion_noise": 0.002,
+            "step_size": 1.0,
+            "iterations": 500,
+            "epsilon": 40.0
+        },
+        "inference_normalization": "off",
+        "use_adaptive_eps": False,
+        "use_adaptive_step": True,
+        "mask_center_x": 0.5,
+        "mask_center_y": 0.0,
+        "mask_radius": 0.2,
+        "mask_sigma": 0.2,
+        "eps_max_mult": 20.0,
+        "eps_min_mult": 1.0,
+        "step_max_mult": 50.0,
+        "step_min_mult": 0.2,
     }
 ]
 
@@ -215,6 +275,10 @@ def run_inference(image, model_type, inference_type, eps_value, num_iterations,
         config['step_size'] = step_size_f
         config['top_layer'] = model_layer
     
+    # Inference normalization off (option removed from UI)
+    config['inference_normalization'] = 'off'
+    config['recognition_normalization'] = 'off'
+    
     # Adaptive epsilon (Gaussian mask)
     if use_adaptive_eps:
         config['adaptive_epsilon'] = {
@@ -323,7 +387,7 @@ def draw_mask_overlay(image, center_x, center_y, radius):
     draw.ellipse(
         [cx_px - radius_px, cy_px - radius_px, cx_px + radius_px, cy_px + radius_px],
         outline="#E11D48",
-        width=max(2, min(w, h) // 150),
+        width=2 * max(2, min(w, h) // 150),
     )
     # Center dot
     r = max(2, min(w, h) // 80)
@@ -331,21 +395,29 @@ def draw_mask_overlay(image, center_x, center_y, radius):
     return img
 
 
-# Helper function to apply example parameters (adaptive mask off by default for examples)
+# Helper function to apply example parameters (adaptive mask off by default unless example defines it)
 def apply_example(example):
+    rd = example["reverse_diff"]
     return [
         example["image"],
-        "resnet50_robust",
+        rd.get("model", "resnet50_robust"),
         example["method"],
-        example["reverse_diff"]["epsilon"],
-        example["reverse_diff"]["iterations"],
-        example["reverse_diff"]["initial_noise"],
-        example["reverse_diff"]["diffusion_noise"],
-        example["reverse_diff"]["step_size"],
-        example["reverse_diff"]["layer"],
-        False, False,  # use_adaptive_eps, use_adaptive_step
-        0.0, 0.0, 0.3, 0.2,  # mask center x/y, radius, sigma
-        4.0, 1.0, 4.0, 1.0,  # eps max/min mult, step max/min mult
+        rd["epsilon"],
+        rd["iterations"],
+        rd["initial_noise"],
+        rd["diffusion_noise"],
+        rd["step_size"],
+        rd["layer"],
+        example.get("use_adaptive_eps", False),
+        example.get("use_adaptive_step", False),
+        example.get("mask_center_x", 0.0),
+        example.get("mask_center_y", 0.0),
+        example.get("mask_radius", 0.3),
+        example.get("mask_sigma", 0.2),
+        example.get("eps_max_mult", 4.0),
+        example.get("eps_min_mult", 1.0),
+        example.get("step_max_mult", 4.0),
+        example.get("step_min_mult", 1.0),
         gr.Group(visible=True),
     ]
 
@@ -360,16 +432,15 @@ with gr.Blocks(title="Human Hallucination Prediction", css="""
     background-color: #7C3AED !important;
 }
 """) as demo:
-    gr.Markdown("# 👁️ Human Hallucination Prediction")
-    gr.Markdown("**Predict what visual hallucinations humans will experience** using adversarially robust neural networks. This demo forecasts perceptual phenomena like illusory contours, figure-ground reversals, and Gestalt effects before humans report them.")
+    gr.Markdown("# Human Hallucination Prediction")
+    gr.Markdown("**Predict what visual hallucinations humans may experience** using neural networks.")
     
     gr.Markdown("""
     **How to predict hallucinations:**
-    1. **Select an example illusion** below and click "Load Parameters" to set optimal prediction settings
+    1. **Select an example image** below and click "Load Parameters" to set the prediction settings
     2. **Click "Run Generative Inference"** to predict what hallucination humans will perceive
     3. **View the prediction**: Watch as the model reveals the perceptual structures it expects—matching what humans typically hallucinate
-    4. **Upload your own images** to test if they will induce hallucinations in human observers
-    """)
+    4. **You can upload your own images** 
     
     # Main processing interface
     with gr.Row():
@@ -403,21 +474,21 @@ with gr.Blocks(title="Human Hallucination Prediction", css="""
                     )
                 
                 with gr.Row():
-                    eps_slider = gr.Slider(minimum=0.0, maximum=40.0, value=10.0, step=0.01, label="Epsilon (Stimulus Fidelity)")
-                    iterations_slider = gr.Slider(minimum=1, maximum=600, value=501, step=1, label="Number of Iterations")
+                    eps_slider = gr.Slider(minimum=0.0, maximum=40.0, value=40.0, step=0.01, label="Epsilon (Stimulus Fidelity)")
+                    iterations_slider = gr.Slider(minimum=1, maximum=600, value=500, step=1, label="Number of Iterations")
                 
                 with gr.Row():
-                    initial_noise_slider = gr.Slider(minimum=0.0, maximum=1.0, value=0.1, step=0.01, 
+                    initial_noise_slider = gr.Slider(minimum=0.0, maximum=1.0, value=1.0, step=0.01, 
                                                    label="Drift Noise")
                     diffusion_noise_slider = gr.Slider(minimum=0.0, maximum=0.05, value=0.002, step=0.001, 
                                                     label="Diffusion Noise")
                     
                 with gr.Row():
-                    step_size_slider = gr.Slider(minimum=0.01, maximum=2.0, value=0.1, step=0.01, 
+                    step_size_slider = gr.Slider(minimum=0.01, maximum=2.0, value=1.0, step=0.01, 
                                                label="Update Rate")
                     layer_choice = gr.Dropdown(
                         choices=["all", "conv1", "bn1", "relu", "maxpool", "layer1", "layer2", "layer3", "layer4", "avgpool"],
-                        value="layer3",
+                        value="all",
                         label="Model Layer"
                     )
                 
@@ -427,17 +498,17 @@ with gr.Blocks(title="Human Hallucination Prediction", css="""
                     use_adaptive_eps_check = gr.Checkbox(value=False, label="Use adaptive epsilon (stronger/weaker constraint by region)")
                     use_adaptive_step_check = gr.Checkbox(value=True, label="Use adaptive step size (stronger/weaker updates by region)")
                 with gr.Row():
-                    mask_center_x_slider = gr.Slider(minimum=-1.0, maximum=1.0, value=0.0, step=0.05, label="Mask center X")
+                    mask_center_x_slider = gr.Slider(minimum=-1.0, maximum=1.0, value=0.5, step=0.05, label="Mask center X")
                     mask_center_y_slider = gr.Slider(minimum=-1.0, maximum=1.0, value=0.0, step=0.05, label="Mask center Y")
                 with gr.Row():
                     mask_radius_slider = gr.Slider(minimum=0.01, maximum=1.0, value=0.2, step=0.01, label="Mask radius (flat region size)")
                     mask_sigma_slider = gr.Slider(minimum=0.05, maximum=0.5, value=0.2, step=0.01, label="Mask sigma (fall-off outside radius)")
                 with gr.Row():
-                    eps_max_mult_slider = gr.Slider(minimum=0.1, maximum=15.0, value=4.0, step=0.1, label="Epsilon: multiplier at center")
+                    eps_max_mult_slider = gr.Slider(minimum=0.1, maximum=350.0, value=20.0, step=0.1, label="Epsilon: multiplier at center")
                     eps_min_mult_slider = gr.Slider(minimum=0.1, maximum=10.0, value=1.0, step=0.1, label="Epsilon: multiplier at periphery")
                 with gr.Row():
-                    step_max_mult_slider = gr.Slider(minimum=0.1, maximum=15.0, value=10.0, step=0.1, label="Step size: multiplier at center")
-                    step_min_mult_slider = gr.Slider(minimum=0.1, maximum=10.0, value=1.0, step=0.1, label="Step size: multiplier at periphery")
+                    step_max_mult_slider = gr.Slider(minimum=0.1, maximum=150.0, value=50.0, step=0.1, label="Step size: multiplier at center")
+                    step_min_mult_slider = gr.Slider(minimum=0.1, maximum=10.0, value=0.2, step=0.1, label="Step size: multiplier at periphery")
             
         with gr.Column(scale=2):
             # Outputs
@@ -445,8 +516,8 @@ with gr.Blocks(title="Human Hallucination Prediction", css="""
             output_frames = gr.Gallery(label="Hallucination Prediction Process", columns=5, rows=2)
     
     # Examples section with integrated explanations
-    gr.Markdown("## 🎯 Visual Illusion Examples")
-    gr.Markdown("Select an illusion to predict what hallucination humans will experience when viewing it")
+    gr.Markdown("## Examples")
+    gr.Markdown("Select an example and click Load Parameters to apply its settings")
     
     # For each example, create a row with the image and explanation side by side
     for i, ex in enumerate(examples):