Allow two usages

app.py

@@ -39,18 +39,40 @@ def classify_preloaded_image(subdirectory, image_name, model_name):
     image = Image.open(image_path).convert("RGB")
     return predict(image, model_name)
 
+def display_selected_image(subdirectory, image_name):
+    """Display the selected image."""
+    image_path = os.path.join(TEST_IMAGE_DIR, subdirectory, image_name)
+    if os.path.exists(image_path):
+        return Image.open(image_path).convert("RGB")
+    return None
+
+def classify_uploaded_image(image, model_name):
+    return predict(image, model_name)
+
 model_choices = list(model_paths.keys())
 
 # Define Gradio app
 with gr.Blocks() as demo:
     gr.Markdown("# Plant Disease Classifier")
+    
+    with gr.Tab("Upload an Image"):
+        with gr.Row():
+            image_input = gr.Image(type="pil", label="Upload an Image")
+            model_input_upload = gr.Dropdown(choices=model_choices, label="Select Model", value="resnet")
+        classify_button_upload = gr.Button("Classify")
+        output_text_upload = gr.Textbox(label="Predicted Class")
+        classify_button_upload.click(classify_uploaded_image, inputs=[image_input, model_input_upload], outputs=output_text_upload)
+
 
     with gr.Tab("Select a Preloaded Image"):
         with gr.Row():
             subdir_dropdown = gr.Dropdown(choices=get_subdirectories(TEST_IMAGE_DIR), label="Select a Subdirectory")
             image_dropdown = gr.Dropdown(choices=[], label="Select an Image")
             model_input_preloaded = gr.Dropdown(choices=model_choices, label="Select Model", value="resnet")
-
+        
+        with gr.Row():
+            image_display = gr.Image(label="Selected Image", interactive=False)
+        
         classify_button_preloaded = gr.Button("Classify")
         output_text_preloaded = gr.Textbox(label="Predicted Class")
 
@@ -59,6 +81,10 @@ with gr.Blocks() as demo:
             return gr.update(choices=get_images_in_subdirectory(subdirectory))
         
         subdir_dropdown.change(update_images, inputs=subdir_dropdown, outputs=image_dropdown)
+
+        # Update displayed image based on selected image
+        image_dropdown.change(display_selected_image, inputs=[subdir_dropdown, image_dropdown], outputs=image_display)
+
         classify_button_preloaded.click(
             classify_preloaded_image, inputs=[subdir_dropdown, image_dropdown, model_input_preloaded], outputs=output_text_preloaded
         )