Update app.py

app.py

@@ -1,83 +1,47 @@
 import gradio as gr
+from transformers import LayoutLMv3ForTokenClassification, LayoutLMv3Processor
+from PIL import Image
 import torch
-from transformers import LayoutLMv3Processor, LayoutLMv3ForTokenClassification
-import pytesseract
-import os
 
-# Explicitly set the Tesseract path for Hugging Face Spaces
-pytesseract.pytesseract.tesseract_cmd = "/usr/bin/tesseract"
+# Load the fine-tuned model and processor
+model_path = "quadranttechnologies/Receipt_Image_Analyzer"  
+model = LayoutLMv3ForTokenClassification.from_pretrained(model_path)
+processor = LayoutLMv3Processor.from_pretrained(model_path)
 
-# Debugging: Print Tesseract version and PATH details
-try:
-    tesseract_version = pytesseract.get_tesseract_version()
-    print("Tesseract Version:", tesseract_version)
-    print("Tesseract Path:", pytesseract.pytesseract.tesseract_cmd)
-    print("Environment PATH:", os.environ["PATH"])
-except Exception as e:
-    print("Tesseract Debugging Error:", e)
+# Define label mapping
+id2label = {0: "company", 1: "date", 2: "address", 3: "total", 4: "other"}
 
-# For local development on Windows
-# Uncomment the line below if running locally on Windows
-# pytesseract.pytesseract.tesseract_cmd = r"C:\Program Files\Tesseract-OCR\tesseract.exe"
+# Define prediction function
+def predict_receipt(image):
+    # Preprocess the image
+    encoding = processor(image, return_tensors="pt", truncation=True, padding="max_length", max_length=512)
+    input_ids = encoding["input_ids"]
+    attention_mask = encoding["attention_mask"]
+    bbox = encoding["bbox"]
+    pixel_values = encoding["pixel_values"]
 
-# Load the model and processor
-processor = LayoutLMv3Processor.from_pretrained("quadranttechnologies/Receipt_Image_Analyzer")
-model = LayoutLMv3ForTokenClassification.from_pretrained("quadranttechnologies/Receipt_Image_Analyzer")
-model.eval()
-device = torch.device("cuda" if torch.cuda.is_available() else "cpu")
-model.to(device)
+    # Get model predictions
+    outputs = model(input_ids=input_ids, attention_mask=attention_mask, bbox=bbox, pixel_values=pixel_values)
+    predictions = outputs.logits.argmax(-1).squeeze().tolist()
 
-def process_image(image):
-    try:
-        # Preprocess the image using the processor
-        encoding = processor(image, return_tensors="pt", truncation=True, padding="max_length", max_length=512)
+    # Map predictions to labels
+    labeled_output = {id2label[pred]: idx for idx, pred in enumerate(predictions) if pred != 4}
 
-        # Move inputs to the same device as the model
-        encoding = {key: val.to(device) for key, val in encoding.items()}
+    return labeled_output
 
-        # Perform inference
-        with torch.no_grad():
-            outputs = model(**encoding)
-            predictions = torch.argmax(outputs.logits, dim=-1)
-
-        # Extract input IDs, bounding boxes, and predicted labels
-        words = encoding["input_ids"]
-        bboxes = encoding["bbox"]
-        labels = predictions.squeeze().tolist()
-
-        # Format output as JSON
-        structured_output = []
-        for word_id, bbox, label in zip(words.squeeze().tolist(), bboxes.squeeze().tolist(), labels):
-            # Decode the word ID to text
-            word = processor.tokenizer.decode([word_id]).strip()
-            if word:  # Avoid adding empty words
-                structured_output.append({
-                    "word": word,
-                    "bounding_box": bbox,
-                    "label": model.config.id2label[label]  # Convert label ID to label name
-                })
-
-        return structured_output
-
-    except Exception as e:
-        # Debugging: Log any errors encountered during processing
-        print("Error during processing:", str(e))
-        return {"error": str(e)}
-
-# Define the Gradio interface
+# Create Gradio Interface
 interface = gr.Interface(
-    fn=process_image,
-    inputs=gr.Image(type="pil"),  # Accepts image input
-    outputs="json",  # Outputs JSON structure
-    title="Receipt_Image_Analyzer",
-    description="Upload an image (e.g., receipt or document) to extract structured information in JSON format."
+    fn=predict_receipt,
+    inputs=gr.inputs.Image(type="pil"),
+    outputs="json",
+    title="Receipt Information Analyzer",
+    description="Upload a scanned receipt image to extract information like company name, date, address, and total."
 )
 
-# Launch the app
+# Launch the interface
 if __name__ == "__main__":
-    # Debugging: Check if the app is starting correctly
-    print("Starting Table OCR App...")
-    interface.launch(share=True)
+    interface.launch()
+