Update app.py

app.py

@@ -52,7 +52,6 @@ for root, dirs, files in os.walk(extracted_path):
 
 # Path to the dataset directory
 data_dir = pathlib.Path('extracted_files/Pest_Dataset')
-data_dir = pathlib.Path(data_dir)
 
 image_count = len(list(data_dir.glob('*/*.jpg')))
 print(image_count)
@@ -60,7 +59,6 @@ print(image_count)
 bees = list(data_dir.glob('bees/*'))
 print(bees[0])
 PIL.Image.open(str(bees[0]))
-
 batch_size = 32
 img_height = 180
 img_width = 180
@@ -73,7 +71,6 @@ train_ds = tf.keras.utils.image_dataset_from_directory(
   image_size=(img_height, img_width),
   batch_size=batch_size)
 
-
 val_ds = tf.keras.utils.image_dataset_from_directory(
   data_dir,
   validation_split=0.2,
@@ -82,7 +79,6 @@ val_ds = tf.keras.utils.image_dataset_from_directory(
   image_size=(img_height, img_width),
   batch_size=batch_size)
 
-
 class_names = train_ds.class_names
 print(class_names)
 
@@ -94,7 +90,6 @@ for images, labels in train_ds.take(1):
     plt.title(class_names[labels[i]])
     plt.axis("off")
 
-
 for image_batch, labels_batch in train_ds:
   print(image_batch.shape)
   print(labels_batch.shape)
@@ -105,7 +100,6 @@ AUTOTUNE = tf.data.AUTOTUNE
 train_ds = train_ds.cache().shuffle(1000).prefetch(buffer_size=AUTOTUNE)
 val_ds = val_ds.cache().prefetch(buffer_size=AUTOTUNE)
 
-
 normalization_layer = layers.Rescaling(1./255)
 
 normalized_ds = train_ds.map(lambda x, y: (normalization_layer(x), y))
@@ -122,6 +116,7 @@ data_augmentation = keras.Sequential(
                                   3)),
     layers.RandomRotation(0.1),
     layers.RandomZoom(0.1),
+    layers.RandomContrast(0.1),
   ]
 )
 
@@ -133,7 +128,6 @@ for images, _ in train_ds.take(1):
     plt.imshow(augmented_images[0].numpy().astype("uint8"))
     plt.axis("off")
 
-
 from tensorflow.keras.applications import EfficientNetB0
 
 base_model = EfficientNetB0(weights='imagenet', include_top=False, input_shape=(img_height, img_width, 3))
@@ -149,7 +143,6 @@ x = keras.layers.GlobalAveragePooling2D()(x)
 x = keras.layers.Dropout(0.2)(x)
 outputs = keras.layers.Dense(len(class_names), activation='softmax')(x)
 
-
 # Compile the model
 model = keras.Model(inputs, outputs)
 model.compile(optimizer='adam',
@@ -190,17 +183,51 @@ plt.legend(loc='upper right')
 plt.title('Training and Validation Loss')
 plt.show()
 
+test_ds = tf.keras.utils.image_dataset_from_directory(
+  data_dir,
+  validation_split=0.2,
+  subset="validation",
+  seed=123,
+  image_size=(img_height, img_width),
+  batch_size=batch_size)
 
-# Evaluate the model on the validation dataset
-results = model.evaluate(val_ds, verbose=0)
+results = model.evaluate(test_ds, verbose=0)
 
-print("Validation Loss: {:.5f}".format(results[0]))
-print("Validation Accuracy: {:.2f}%".format(results[1] * 100))
+print("    Test Loss: {:.5f}".format(results[0]))
+print("Test Accuracy: {:.2f}%".format(results[1] * 100))
 
 
-import gradio as gr
-import numpy as np
-import tensorflow as tf
+# Metrics
+y_true = []
+y_pred = []
+
+for images, labels in test_ds:
+    y_true.extend(labels.numpy())
+    preds = model.predict(images)
+    y_pred.extend(np.argmax(preds, axis=1))
+
+from sklearn.metrics import classification_report, confusion_matrix
+print(classification_report(y_true, y_pred, target_names=class_names))
+
+import pandas as pd
+report = classification_report(y_true, y_pred, target_names=class_names, output_dict=True)
+df = pd.DataFrame(report).transpose()
+print(df)
+
+def make_confusion_matrix(y_true, y_pred, labels):
+    cm = confusion_matrix(y_true, y_pred)
+    fig, ax = plt.subplots(figsize=(10, 8))
+    cax = ax.matshow(cm, cmap=plt.cm.Blues)
+    plt.title('Confusion Matrix')
+    fig.colorbar(cax)
+    ax.set_xticklabels([''] + labels, rotation=90)
+    ax.set_yticklabels([''] + labels)
+    plt.xlabel('Predicted')
+    plt.ylabel('True')
+    plt.show()
+
+make_confusion_matrix(y_true, y_pred, class_names)
+
 
 def predict_image(img):
     img = np.array(img)
@@ -209,9 +236,8 @@ def predict_image(img):
     prediction = model.predict(img_4d)[0]
     return {class_names[i]: float(prediction[i]) for i in range(len(class_names))}
 
-image = gr.Image()
-label = gr.Label(num_top_classes=1)
-
+image = gr.Image(type="pil")
+label = gr.Label(num_top_classes=12)
 
 # Define custom CSS for background image
 custom_css = """