Update app.py

app.py

@@ -1,26 +1,26 @@
-import numpy as np 
+from transformers import DetrImageProcessor, DetrForObjectDetection
+import torch
 from PIL import Image
-from transformers import AutoImageProcessor, AutoModelForImageClassification
+import requests
 
-# Load the model and image processor
-processor = AutoImageProcessor.from_pretrained("beingamit99/car_damage_detection")
-model = AutoModelForImageClassification.from_pretrained("beingamit99/car_damage_detection")
+url = "http://images.cocodataset.org/val2017/000000039769.jpg"
+image = Image.open(requests.get(url, stream=True).raw)
 
-# Load and process the image
-image = Image.open(IMAGE)
-inputs = processor(images=image, return_tensors="pt")
+# you can specify the revision tag if you don't want the timm dependency
+processor = DetrImageProcessor.from_pretrained("facebook/detr-resnet-50", revision="no_timm")
+model = DetrForObjectDetection.from_pretrained("facebook/detr-resnet-50", revision="no_timm")
 
-# Make predictions
+inputs = processor(images=image, return_tensors="pt")
 outputs = model(**inputs)
-logits = outputs.logits.detach().cpu().numpy()
-predicted_class_id = np.argmax(logits)
-predicted_proba = np.max(logits)
-label_map = model.config.id2label
-predicted_class_name = label_map[predicted_class_id]
 
-# Print the results
-print(f"Predicted class: {predicted_class_name} (probability: {predicted_proba:.4f}")
-from transformers import pipeline
-#Create a classification pipeline
-pipe = pipeline("image-classification", model="beingamit99/car_damage_detection")
-pipe(IMAGE)
+# convert outputs (bounding boxes and class logits) to COCO API
+# let's only keep detections with score > 0.9
+target_sizes = torch.tensor([image.size[::-1]])
+results = processor.post_process_object_detection(outputs, target_sizes=target_sizes, threshold=0.9)[0]
+
+for score, label, box in zip(results["scores"], results["labels"], results["boxes"]):
+    box = [round(i, 2) for i in box.tolist()]
+    print(
+            f"Detected {model.config.id2label[label.item()]} with confidence "
+            f"{round(score.item(), 3)} at location {box}"
+    )