Update app.py

app.py

@@ -4,17 +4,22 @@ model = torch.hub.load('pytorch/vision:v0.6.0', 'resnet18', pretrained=True).eva
 
 
 import requests
-from PIL import Image
+import PIL
 from torchvision import transforms
 
 # Download human-readable labels for ImageNet.
 response = requests.get("https://git.io/JJkYN")
-labels = response.text.split("\n")
-
-def predict(inp):
-  inp = transforms.ToTensor()(inp).unsqueeze(0)
+labels   = response.text.split("\n")
+
+def classify_image(image_filepath):
+  PIL_image     = PIL.Image.open(image_filepath).convert('RGB')
+  transformations = transforms.Compose([
+    transforms.Resize(size = (224,224)),
+    transforms.ToTensor(),
+  ])
+  image_tensors = transformations(PIL_image).unsqueeze(0)
   with torch.no_grad():
-    prediction = torch.nn.functional.softmax(model(inp)[0], dim=0)
+    prediction = torch.nn.functional.softmax(model(image_tensors)[0], dim=0)
     confidences = {labels[i]: float(prediction[i]) for i in range(1000)}
   return confidences
 
@@ -29,7 +34,7 @@ with gr.Blocks(title="Image Classification for 1000 Objects", css=".gradio-conta
         output_label = gr.Label(label="Probabilities", num_top_classes=3)
     
     send_btn = gr.Button("Infer")
-    send_btn.click(fn=predict, inputs=input_image, outputs=output_label)
+    send_btn.click(fn=classify_image, inputs=input_image, outputs=output_label)
     
     with gr.Row():
         gr.Examples(['./lion.jpg']   , label='Sample images : Lion', inputs=input_image)