Update app.py

app.py

@@ -4,7 +4,10 @@ import gradio as gr
 from transformers import AutoTokenizer, AutoFeatureExtractor, VisionEncoderDecoderModel
 from PIL import Image
 import torch
-
+import os
+os.system('pip install paddlepaddle')
+os.system('pip install paddleocr')
+from paddleocr import PaddleOCR, draw_ocr
 
 tokenizer = AutoTokenizer.from_pretrained("kha-white/manga-ocr-base")
 
@@ -22,7 +25,10 @@ def post_process(text):
   return text
 
 def manga_ocr(img):
-  img = img.convert('L').convert('RGB')
+  ocr = PaddleOCR(use_angle_cls=True, lang='japan',use_gpu=False)
+  img_path = img.name
+  result = ocr.ocr(img_path, cls=True)
+  image = Image.open(img_path).convert('RGB')
   pixel_values = feature_extractor(img, return_tensors="pt").pixel_values
   output = model.generate(pixel_values)[0]
   text = tokenizer.decode(output, skip_special_tokens=True)