gradio app

app.py

@@ -0,0 +1,29 @@
+import gradio as gr
+from PIL import Image
+import torchvision
+import torch
+
+# load model
+
+def predict(input_image):
+    pil_image = Image.fromarray(input_image.astype('uint8'), 'RGB')
+    # transform image to torch and do preprocessing
+    torch_image = torchvision.transforms.ToTensor()(pil_image)
+    # model predict
+    prediction = torch.rand(torch_image.shape)
+    # transform torch to image
+    predicted_pil_image = torchvision.transforms.ToPILImage()(prediction)
+    # return correct image
+    return predicted_pil_image
+
+iface = gr.Interface(
+    fn=predict, 
+    inputs=gr.Image(shape=(512,512)), 
+    outputs=gr.Image(shape=(512,512)),
+    examples=[
+        ["demo_imgs/fake.jpg"] # use real image
+    ],
+    title="DTM Estimation",
+    description="This demo predict a DTM..."
+)
+iface.launch()

requirements.txt

@@ -0,0 +1,3 @@
+gradio
+torch
+torchvision