Spaces:
Sleeping
Sleeping
| import torch | |
| import numpy as np | |
| from transformers import LayoutLMv3TokenizerFast, LayoutLMv3Processor, LayoutLMv3ForTokenClassification | |
| from PIL import Image, ImageDraw, ImageFont | |
| from utils import OCR, unnormalize_box | |
| tokenizer = LayoutLMv3TokenizerFast.from_pretrained("mp-02/layoutlmv3-base-cord-sroie", apply_ocr=False) | |
| processor = LayoutLMv3Processor.from_pretrained("mp-02/layoutlmv3-base-cord-sroie", apply_ocr=False) | |
| model = LayoutLMv3ForTokenClassification.from_pretrained("mp-02/layoutlmv3-base-cord-sroie") | |
| id2label = model.config.id2label | |
| label2id = model.config.label2id | |
| device = torch.device('cuda' if torch.cuda.is_available() else 'cpu') | |
| model.to(device) | |
| import json | |
| def prediction(image): | |
| boxes, words = OCR(image) | |
| # Preprocessa l'immagine e il testo con il processore di LayoutLMv3 | |
| encoding = processor(image, words, boxes=boxes, return_offsets_mapping=True, return_tensors="pt", truncation=True) | |
| offset_mapping = encoding.pop('offset_mapping') | |
| # Esegui l'inferenza con il modello fine-tuned | |
| outputs = model(**encoding) | |
| predictions = outputs.logits.argmax(-1).squeeze().tolist() | |
| token_boxes = encoding.bbox.squeeze().tolist() | |
| probabilities = torch.softmax(outputs.logits, dim=-1) | |
| confidence_scores = probabilities.max(-1).values.squeeze().tolist() | |
| inp_ids = encoding.input_ids.squeeze().tolist() | |
| inp_words = [tokenizer.decode(i) for i in inp_ids] | |
| width, height = image.size | |
| is_subword = np.array(offset_mapping.squeeze().tolist())[:, 0] != 0 | |
| true_predictions = [id2label[pred] for idx, pred in enumerate(predictions) if not is_subword[idx]] | |
| true_boxes = [unnormalize_box(box, width, height) for idx, box in enumerate(token_boxes) if not is_subword[idx]] | |
| true_confidence_scores = [confidence_scores[idx] for idx, conf in enumerate(confidence_scores) if not is_subword[idx]] | |
| true_words = [] | |
| for id, i in enumerate(inp_words): | |
| if not is_subword[id]: | |
| true_words.append(i) | |
| else: | |
| true_words[-1] = true_words[-1]+i | |
| true_predictions = true_predictions[1:-1] | |
| true_boxes = true_boxes[1:-1] | |
| true_words = true_words[1:-1] | |
| true_confidence_scores = true_confidence_scores[1:-1] | |
| d = {} | |
| for id, i in enumerate(true_predictions): | |
| #rimuovo i prefissi | |
| if i != "O": | |
| i = i[2:] | |
| if i not in d.keys(): | |
| d[i] = true_words[id] | |
| else: | |
| d[i] = d[i] + ", " + true_words[id] | |
| d = {k: v.strip() for (k, v) in d.items()} | |
| if "O" in d: d.pop("O") | |
| if("MENU.NM" in d and "MENU.PRICE" in d): | |
| if(len(d["MENU.NM"].split(", ")) == len(d["MENU.PRICE"].split(", "))): | |
| menu_names = [name.strip() for name in d["MENU.NM"].split(', ')] | |
| menu_prices = [price.strip() for price in d["MENU.PRICE"].split(', ')] | |
| menu_combined = [{"ITEM": name, "PRICE": price} for name, price in zip(menu_names, menu_prices)] | |
| d.pop("MENU.NM") | |
| d.pop("MENU.PRICE") | |
| d["MENU"] = menu_combined | |
| draw = ImageDraw.Draw(image, "RGBA") | |
| font = ImageFont.load_default() | |
| for prediction, box, confidence in zip(true_predictions, true_boxes, true_confidence_scores): | |
| draw.rectangle(box) | |
| draw.text((box[0]+10, box[1]-10), text=str(prediction)+ ", "+ str(confidence), font=font, fill="black", font_size="15") | |
| return d | |