Update app.py

app.py

@@ -1,29 +1,23 @@
 import gradio as gr
 from PIL import Image
-from transformers import AutoProcessor, VisionEncoderDecoderModel
-import torch
+from doctr.models import ocr_predictor
+from doctr.io import DocumentFile
 
-# Load Parseq model
-model_name = "naver-clova-ix/parseq"
-processor = AutoProcessor.from_pretrained(model_name)
-model = VisionEncoderDecoderModel.from_pretrained(model_name)
-device = torch.device("cuda" if torch.cuda.is_available() else "cpu")
-model.to(device)
+# Load the OCR model
+model = ocr_predictor(det_arch='db_resnet50', reco_arch='parseq', pretrained=True)
 
 def recognize_handwriting(image):
     if image is None:
         return "Please upload an image."
 
-    image = image.convert("RGB")
-    inputs = processor(images=image, return_tensors="pt").to(device)
-    outputs = model.generate(**inputs)
-    decoded = processor.batch_decode(outputs, skip_special_tokens=True)[0]
-
-    return decoded.strip()
+    doc = DocumentFile.from_images(image)
+    result = model(doc)
+    text = "\n".join([block.value for page in result.pages for block in page.blocks])
+    return text.strip()
 
 with gr.Blocks() as demo:
     gr.Markdown("![Logo](logo.png)")
-    gr.Markdown("Evaluator! Imrove Your Writing")
+    gr.Markdown("# 📝 Handwriting Recognition with Doctr")
 
     with gr.Row():
         image_input = gr.Image(type="pil", label="Upload Handwritten Image")
@@ -32,5 +26,4 @@ with gr.Blocks() as demo:
     extract_btn = gr.Button("Extract Text")
     extract_btn.click(fn=recognize_handwriting, inputs=image_input, outputs=text_output)
 
-if __name__ == "__main__":
-    demo.launch()
+demo.launch()