Update app.py

app.py

@@ -1,96 +1,96 @@
-import tensorflow as tf
-from keras.models import model_from_json
-import requests
-import gradio as gr
-import numpy as np
-import PIL.Image as Image
-from pathlib import Path
-from glob import glob
-import cv2
-import os
-
-dataset = 'Color'
-
-model_architecture =  'Models/Model_Color_128_128.json'
-
-model_weights = 'Models/Model_Color_128_128.h5'
-
-RESCALE_IMG_WIDTH, RESCALE_IMG_HEIGHT = 128, 128
-SIZE = 128
-
-
-def resize_image_pil(img, new_width, new_height):
-
-    # Convert to PIL image
-    img = Image.fromarray(img)
-    
-    # Get original size
-    width, height = img.size
-
-    # Calculate scale
-    width_scale = new_width / width
-    height_scale = new_height / height 
-    scale = min(width_scale, height_scale)
-
-    # Resize
-    resized = img.resize((int(width*scale), int(height*scale)), Image.NEAREST)
-    
-    # Crop to exact size
-    resized = resized.crop((0, 0, new_width, new_height))
-
-    return resized
-
-
-# def transform_img(img):
-#     img_blur = cv2.GaussianBlur(img, (3 ,3), 0, 0)
-#     kernel = np.array([[-3,-1,-1], [-1,9,-1], [-1,-1,1]])
-#     im = cv2.filter2D(img_blur, -1, kernel)
-#     img_inv = cv2.bitwise_not(im)
-
-    return img_inv
-
-def classify_image(inp):
-    label = ['Dog','Cat']
-
-    # inp2 = transform_img(inp)
-    img_resized = resize_image_pil(inp,RESCALE_IMG_WIDTH, RESCALE_IMG_HEIGHT) 
-    img = np.array(img_resized)
-    img = img * (1/255)
-    reshaped_array = img.reshape((-1, RESCALE_IMG_WIDTH, RESCALE_IMG_HEIGHT , 3))
-
-    predictions = model.predict(x=reshaped_array).flatten()
-
-    confidences = {label[i]: float(predictions[i]) for i in range(2)}
-
-    # img_resized = transform_img(inp)
-    img_resized = Image.fromarray(img_resized)
-    img_resized = img_resized.resize(size = (512,512), resample = Image.NEAREST)
-
-    return [confidences, img_resized]
-
-
-# load json and create model
-json_file = open(model_architecture, 'r')
-loaded_model_json = json_file.read()
-json_file.close()
-
-model = model_from_json(loaded_model_json)
-# load weights into new model
-model.load_weights(model_weights)
-
-np.random.seed(123)
-random_imgs = np.random.choice(
-    glob( dataset + '/**/*.jpg'),
-    size=128,
-    replace=False,
-)
-
-demo = gr.Interface(
-    fn=classify_image,
-    inputs=gr.Image(width=256, height=256),
-    outputs=[gr.Label(num_top_classes=2), gr.Image(width=512, height=512)],
-    examples=[img_path for img_path in random_imgs],
-    examples_per_page=64,
-    title= 'Cat & Dog classifier'
-)
-demo.launch(inbrowser=True)
+import tensorflow as tf
+from keras.models import model_from_json
+import requests
+import gradio as gr
+import numpy as np
+import PIL.Image as Image
+from pathlib import Path
+from glob import glob
+import cv2
+import os
+
+dataset = 'Color'
+
+model_architecture =  'Models/Model_Color_128_128.json'
+
+model_weights = 'Models/Model_Color_128_128.h5'
+
+RESCALE_IMG_WIDTH, RESCALE_IMG_HEIGHT = 128, 128
+SIZE = 128
+
+
+def resize_image_pil(img, new_width, new_height):
+
+    # Convert to PIL image
+    img = Image.fromarray(img)
+    
+    # Get original size
+    width, height = img.size
+
+    # Calculate scale
+    width_scale = new_width / width
+    height_scale = new_height / height 
+    scale = min(width_scale, height_scale)
+
+    # Resize
+    resized = img.resize((int(width*scale), int(height*scale)), Image.NEAREST)
+    
+    # Crop to exact size
+    resized = resized.crop((0, 0, new_width, new_height))
+
+    return resized
+
+
+# def transform_img(img):
+#     img_blur = cv2.GaussianBlur(img, (3 ,3), 0, 0)
+#     kernel = np.array([[-3,-1,-1], [-1,9,-1], [-1,-1,1]])
+#     im = cv2.filter2D(img_blur, -1, kernel)
+#     img_inv = cv2.bitwise_not(im)
+
+    return img_inv
+
+def classify_image(inp):
+    label = ['Dog','Cat']
+
+    # inp2 = transform_img(inp)
+    img_resized = resize_image_pil(inp,RESCALE_IMG_WIDTH, RESCALE_IMG_HEIGHT) 
+    img = np.array(img_resized)
+    img = img * (1/255)
+    reshaped_array = img.reshape((-1, RESCALE_IMG_WIDTH, RESCALE_IMG_HEIGHT , 3))
+
+    predictions = model.predict(x=reshaped_array).flatten()
+
+    confidences = {label[i]: float(predictions[i]) for i in range(2)}
+
+    # img_resized = transform_img(inp)
+    img_resized = Image.fromarray(inp)
+    img_resized = img_resized.resize(size = (512,512), resample = Image.NEAREST)
+
+    return [confidences, img_resized]
+
+
+# load json and create model
+json_file = open(model_architecture, 'r')
+loaded_model_json = json_file.read()
+json_file.close()
+
+model = model_from_json(loaded_model_json)
+# load weights into new model
+model.load_weights(model_weights)
+
+np.random.seed(123)
+random_imgs = np.random.choice(
+    glob( dataset + '/**/*.jpg'),
+    size=128,
+    replace=False,
+)
+
+demo = gr.Interface(
+    fn=classify_image,
+    inputs=gr.Image(width=256, height=256),
+    outputs=[gr.Label(num_top_classes=2), gr.Image(width=512, height=512)],
+    examples=[img_path for img_path in random_imgs],
+    examples_per_page=64,
+    title= 'Cat & Dog classifier'
+)
+demo.launch(inbrowser=True)