Added Xai

app.py

@@ -1,7 +1,6 @@
 # %%
 import gradio as gr
 import tensorflow as tf
-import numpy as np
 import cv2
 import os
 
@@ -18,12 +17,25 @@ if not os.path.exists(destination):
         print('Repository cloned successfully.')
     except subprocess.CalledProcessError as e:
         print(f'Error cloning repository: {e.output.decode()}')
+
+if not os.path.exists(destination):
+    import subprocess
+    repo_url = os.getenv("GIT_CORE")
+    command = f'git clone {repo_url}'
+    try:
+        subprocess.check_output(command, stderr=subprocess.STDOUT, shell=True)#, env=env)
+        print('Repository cloned successfully.')
+    except subprocess.CalledProcessError as e:
+        print(f'Error cloning repository: {e.output.decode()}')
+
+from explainer_tf_mobilenetv2.explainer import explainer
 # %%
 with open(f'{model_folder}/labels.txt', 'r') as f:
   labels = f.read().split('\n')
 
 # model = tf.saved_model.load(f'{model_folder}/last_layer.hdf5')
-model = tf.keras.models.load_model(f'{model_folder}/last_layer.hdf5')
+# model = tf.keras.models.load_model(f'{model_folder}/last_layer.hdf5')
+model = tf.keras.models.load_model(f'{model_folder}/MobileNetV2_last_layer.hdf5')
 # %%
 def classify_image(inp):
   inp = cv2.resize(inp, (224,224,))
@@ -34,7 +46,30 @@ def classify_image(inp):
   confidences = {labels[i]: float(prediction[i]) for i in range(len(labels))}
   return confidences
 
-gr.Interface(fn=classify_image, 
-             inputs=gr.Image(shape=(224, 224)),
-             outputs=gr.Label(num_top_classes=3),
-             examples=["TomatoHealthy2.jpg", "TomatoYellowCurlVirus3.jpg", "AppleCedarRust3.jpg"]).launch()
+def explainer_wrapper(inp):
+  return explainer(inp, model)
+
+with gr.Blocks() as demo:
+  with gr.Column():
+    with gr.Row():
+        with gr.Column():
+          image = gr.inputs.Image(shape=(224, 224))
+          with gr.Row():
+            classify = gr.Button("Classify")
+            interpret = gr.Button("Interpret")
+
+        label = gr.outputs.Label(num_top_classes=3)
+        interpretation = gr.Plot(label="Interpretation")
+        # interpretation = gr.outputs.Image(type="numpy", label="Interpretation")
+  gr.Examples(["TomatoHealthy2.jpg", "TomatoYellowCurlVirus3.jpg", "AppleCedarRust3.jpg"],
+              inputs=[image],)
+  classify.click(classify_image, image, label, queue=True)
+  interpret.click(explainer_wrapper, image, interpretation, queue=True)
+  
+
+demo.queue(concurrency_count=3).launch()         
+#%%
+# gr.Interface(fn=classify_image, 
+#              inputs=gr.Image(shape=(224, 224)),
+#              outputs=gr.Label(num_top_classes=3),
+#              examples=["TomatoHealthy2.jpg", "TomatoYellowCurlVirus3.jpg", "AppleCedarRust3.jpg"]).launch()