update app.py

app.py

@@ -54,7 +54,12 @@ def predict(image):
     x_max = int(bbox[2] * width)
     y_max = int(bbox[3] * height)
     print(f"Prediction: {classification} with confidence {confidence:.2f}")
-    return classification, confidence, (x_min, y_min, x_max, y_max), draw_bounding_box(image, (x_min, y_min, x_max, y_max), classification)
+    return (
+        classification, 
+        confidence, 
+        (x_min, y_min, x_max, y_max), 
+        draw_bounding_box(image, (x_min, y_min, x_max, y_max), classification)
+    )
 
 # Function to draw bounding box on the image
 def draw_bounding_box(image, bbox, label):
@@ -67,24 +72,6 @@ def draw_bounding_box(image, bbox, label):
     draw.text(text_position, label, fill="red")
     return image
 
-# Combined predict function that accepts two inputs: an uploaded image and a selected example image.
-def combined_predict(uploaded_image, selected_image_path):
-    print("Combined predict function called.")
-    if uploaded_image is not None:
-        print("Using uploaded image.")
-        image = uploaded_image
-    elif selected_image_path is not None and selected_image_path != "":
-        print("Using selected example image:", selected_image_path)
-        try:
-            image = Image.open(selected_image_path)
-        except Exception as e:
-            print("Error opening selected image:", e)
-            return "Error loading image", "", "", None
-    else:
-        print("No image provided.")
-        return "No image provided", "", "", None
-    return predict(image)
-
 # List of example image file paths (ensure these images are uploaded to your Space)
 example_image_paths = [
     "example1.png",
@@ -93,29 +80,38 @@ example_image_paths = [
     "example4.png"
 ]
 
-# Build the Gradio interface
+# Build the Gradio interface using Blocks
 with gr.Blocks() as demo:
     gr.Markdown("# Seamount Detection")
-    gr.Markdown("**Instructions:** Either upload your own image (left) **or** select one of the example images (right). If both are provided, the uploaded image takes precedence.")
-    
+    gr.Markdown("**Instructions:** Either upload your own image or select one of the example images below. Clicking an example will populate the input field.")
+
     with gr.Row():
-        uploaded_image = gr.Image(type="pil", label="Upload Your Image")
-        selected_image = gr.Radio(choices=example_image_paths, label="Or select an example image", type="value")
+        # Single image input component
+        uploaded_image = gr.Image(type="pil", label="Upload an Image")
+    
+    # Examples component displays the images as clickable thumbnails.
+    gr.Examples(
+        examples=[[path] for path in example_image_paths],
+        inputs=uploaded_image,
+        label="Or select an example image"
+    )
     
     submit_btn = gr.Button("Predict")
     
-    output_class = gr.Text(label="Classification")
-    output_confidence = gr.Text(label="Confidence Score")
-    output_bbox = gr.Text(label="Bounding Box (x_min, y_min, x_max, y_max)")
+    with gr.Row():
+        output_class = gr.Text(label="Classification")
+        output_confidence = gr.Text(label="Confidence Score")
+        output_bbox = gr.Text(label="Bounding Box (x_min, y_min, x_max, y_max)")
+    
     output_image = gr.Image(label="Image with Bounding Box")
     
     submit_btn.click(
-        combined_predict,
-        inputs=[uploaded_image, selected_image],
+        predict,
+        inputs=uploaded_image,
         outputs=[output_class, output_confidence, output_bbox, output_image]
     )
 
-# DEBUG: Launch the app and print a debug message.
+# DEBUG: Launch the app
 print("Launching app...")
 if __name__ == "__main__":
     demo.launch()