Update app.py

app.py

@@ -1,22 +1,45 @@
-import os, io
 import gradio as gr
-from transformers import pipeline
+from transformers import DPTFeatureExtractor, DPTForDepthEstimation
+import torch
+import numpy as np
+from PIL import Image
 
 
-SECRET_TOKEN = os.getenv("SECRET_TOKEN")
-headers = {"Authorization": f'Bearer {SECRET_TOKEN}'}
+feature_extractor = DPTFeatureExtractor.from_pretrained("Intel/dpt-large")
+model = DPTForDepthEstimation.from_pretrained("Intel/dpt-large")
 
-estimator = pipeline("depth-estimation")
+def process_image(image):
+    # prepare image for the model
+    encoding = feature_extractor(image, return_tensors="pt")
+    
+    # forward pass
+    with torch.no_grad():
+       outputs = model(**encoding)
+       predicted_depth = outputs.predicted_depth
+    
+    # interpolate to original size
+    prediction = torch.nn.functional.interpolate(
+                        predicted_depth.unsqueeze(1),
+                        size=image.size[::-1],
+                        mode="bicubic",
+                        align_corners=False,
+                 ).squeeze()
+    output = prediction.cpu().numpy()
+    formatted = (output * 255 / np.max(output)).astype('uint8')
+    img = Image.fromarray(formatted)
+    return img
+    
+    return result
+    
+title = "Depth Estimation"
+description = "Upload an image and get the depth visualization"
+examples =[["house.jpg"], ["plane.webp"], ["room.webp"]]
 
-def rb(img):
-    result = estimator(img)
-    return result["depth"]
-
-description = """Upload an image and get the depth visualization"""
-title = """Depth Estimation"""
-examples=[["house.jpg"], ["plane.webp"], ["room.webp"]]
-inputs = gr.inputs.Image(type="pil", label="Upload an image")
-outputs = gr.outputs.Image(type="pil",label="Output Image")
-
-demo = gr.Interface(fn=rb, inputs=inputs, outputs=outputs, examples=examples, title=title, description=description)
-demo.launch()
+iface = gr.Interface(fn=process_image, 
+                     inputs=gr.inputs.Image(type="pil", label="Upload an image"), 
+                     outputs=gr.outputs.Image(type="pil", label="Predicted depth"),
+                     title=title,
+                     description=description,
+                     examples=examples,
+                     enable_queue=True)
+iface.launch(debug=True)