combining model and upload image code

app.py

@@ -24,22 +24,15 @@ from image_posterior import create_gif
 def get_image_data(inp_image):
     """Gets the image data and model."""
     image = get_dataset_by_name(inp_image, get_label=False)
-    print("image returned\n", image)
+    # print("image returned\n", image)
     model_and_data = process_imagenet_get_model(image)
-    print("model returned\n", model_and_data)
+    # print("model returned\n", model_and_data)
     return image, model_and_data
 
 
 def segmentation_generation(image_name, c_width, n_top, n_gif_imgs):
-    print("Inputs Received:", image_name, c_width, n_top, n_gif_imgs)
+    print("Inputs Received:", image_name, c_width, n_top, n_gif_imgs)    
 
-    get_image_data(image_name)
-
-    return "yeehaw"
-
-    # cred_width = c_width
-    # n_top_segs = n_top
-    # n_gif_images = n_gif_imgs
     image, model_and_data = get_image_data(image_name)
     # print("model_and_data", model_and_data)
     # Unpack datax
@@ -50,10 +43,10 @@ def segmentation_generation(image_name, c_width, n_top, n_gif_imgs):
     label = model_and_data["label"]
     
 
-    if (image_name == 'imagenet_diego'):
-        label = 156
-    elif (image_name == 'imagenet_french_bulldog'):
-        label = 245
+    # if (image_name == 'imagenet_diego'):
+    #     label = 156
+    # elif (image_name == 'imagenet_french_bulldog'):
+    #     label = 245
 
     # Unpack instance and segments
     instance = xtest[0]
@@ -66,8 +59,8 @@ def segmentation_generation(image_name, c_width, n_top, n_gif_imgs):
     xtrain = get_xtrain(segments)
 
     prediction = np.argmax(cur_model(xtrain[:1]), axis=1)
-    if image_name in ["imagenet_diego", "imagenet_french_bulldog"]:
-        assert prediction == label, f"Prediction is {prediction} not {label}"
+    # if image_name in ["imagenet_diego", "imagenet_french_bulldog"]:
+    #     assert prediction == label, f"Prediction is {prediction} not {label}"
 
     # Compute explanation
     exp_init = BayesLocalExplanations(training_data=xtrain,
@@ -85,16 +78,11 @@ def segmentation_generation(image_name, c_width, n_top, n_gif_imgs):
     # Create the gif of the explanation
     return create_gif(rout['blr'], image_name, segments, instance, prediction[0], n_gif_imgs, n_top)
 
-def image_mod(image):
-    return image.rotate(45)
-    
 if __name__ == "__main__":
     # gradio's image inputs look like this: <PIL.Image.Image image mode=RGB size=305x266 at 0x7F3D01C91FA0>
     # need to learn how to handle image inputs, or deal with file inputs or just file path strings
-    # inp = gr.inputs.Textbox(lines=1, placeholder="Select an example from below", default="", label="Input Image Path", optional=False)
-    inp = gr.inputs.Image(label="Input Image", type="pil")
-    out = 'text'
-    # out = [gr.outputs.HTML(label="Output GIF"), gr.outputs.Textbox(label="Prediction")]
+    inp = gr.inputs.Image(label="Input Image (Or select an example)", type="pil")
+    out = [gr.outputs.HTML(label="Output GIF"), gr.outputs.Textbox(label="Prediction")]
 
     iface = gr.Interface(
         segmentation_generation, 
@@ -105,7 +93,7 @@ if __name__ == "__main__":
             gr.inputs.Slider(minimum=10, maximum=50, step=1, default=20, label="n_gif_images", optional=False), 
         ], 
         outputs=out, 
-        examples=[["imagenet_diego", 0.01, 7, 50], 
-                  ["imagenet_french_bulldog", 0.05, 5, 50]]
+        examples=[["./imagenet_diego.png", 0.01, 7, 50], 
+                  ["./imagenet_french_bulldog.jpg", 0.05, 5, 50]]
     )
     iface.launch(enable_queue=True)