Spaces:
Runtime error
Runtime error
File size: 5,649 Bytes
cecb26a 7a45b61 cecb26a 0ed5d15 cecb26a ef6f1b6 cecb26a 0ed5d15 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 |
import os
os.system('pip install pip --upgrade')
os.system('pip install -q git+https://github.com/huggingface/transformers.git')
os.system("pip install pyyaml==5.1")
# workaround: install old version of pytorch since detectron2 hasn't released packages for pytorch 1.9 (issue: https://github.com/facebookresearch/detectron2/issues/3158)
os.system(
"pip install torch==1.8.0+cu101 torchvision==0.9.0+cu101 -f https://download.pytorch.org/whl/torch_stable.html"
)
# install detectron2 that matches pytorch 1.8
# See https://detectron2.readthedocs.io/tutorials/install.html for instructions
os.system(
"pip install -q detectron2 -f https://dl.fbaipublicfiles.com/detectron2/wheels/cu101/torch1.8/index.html"
)
## install PyTesseract
os.system("pip install -q pytesseract")
import gradio as gr
import numpy as np
from transformers import AutoModelForTokenClassification
from datasets.features import ClassLabel
from transformers import AutoProcessor
from datasets import Features, Sequence, ClassLabel, Value, Array2D, Array3D
import torch
from datasets import load_metric
from transformers import LayoutLMv3ForTokenClassification
from transformers.data.data_collator import default_data_collator
from transformers import AutoModelForTokenClassification
from datasets import load_dataset
from PIL import Image, ImageDraw, ImageFont
processor = AutoProcessor.from_pretrained("jinhybr/OCR-LayoutLMv3-Invoice", apply_ocr=True)
model = AutoModelForTokenClassification.from_pretrained("jinhybr/OCR-LayoutLMv3-Invoice")
# load image example
dataset = load_dataset("jinhybr/WildReceipt", split="test")
Image.open(dataset[1]["image_path"]).convert("RGB").save("example1.png")
Image.open(dataset[3]["image_path"]).convert("RGB").save("example2.png")
Image.open(dataset[25]["image_path"]).convert("RGB").save("example3.png")
# define id2label, label2color
labels = dataset.features['ner_tags'].feature.names
id2label = {v: k for v, k in enumerate(labels)}
label2color = {
"Date_key": 'red',
"Date_value": 'green',
"Ignore": 'orange',
"Others": 'orange',
"Prod_item_key": 'red',
"Prod_item_value": 'green',
"Prod_price_key": 'red',
"Prod_price_value": 'green',
"Prod_quantity_key": 'red',
"Prod_quantity_value": 'green',
"Store_addr_key": 'red',
"Store_addr_value": 'green',
"Store_name_key": 'red',
"Store_name_value": 'green',
"Subtotal_key": 'red',
"Subtotal_value": 'green',
"Tax_key": 'red',
"Tax_value": 'green',
"Tel_key": 'red',
"Tel_value": 'green',
"Time_key": 'red',
"Time_value": 'green',
"Tips_key": 'red',
"Tips_value": 'green',
"Total_key": 'red',
"Total_value": 'blue'
}
def unnormalize_box(bbox, width, height):
return [
width * (bbox[0] / 1000),
height * (bbox[1] / 1000),
width * (bbox[2] / 1000),
height * (bbox[3] / 1000),
]
def iob_to_label(label):
return label
def process_image(image):
print(type(image))
width, height = image.size
# encode
encoding = processor(image, truncation=True, return_offsets_mapping=True, return_tensors="pt")
offset_mapping = encoding.pop('offset_mapping')
# forward pass
outputs = model(**encoding)
# get predictions
predictions = outputs.logits.argmax(-1).squeeze().tolist()
token_boxes = encoding.bbox.squeeze().tolist()
# only keep non-subword predictions
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]]
# draw predictions over the image
draw = ImageDraw.Draw(image)
font = ImageFont.load_default()
for prediction, box in zip(true_predictions, true_boxes):
predicted_label = iob_to_label(prediction)
draw.rectangle(box, outline=label2color[predicted_label])
draw.text((box[0]+10, box[1]-10), text=predicted_label, fill=label2color[predicted_label], font=font)
return image
title = "OCR Invoice - Information Extraction - LayoutLMv3"
description = "Fine-tuned Microsoft's LayoutLMv3 on WildReceipt Dataset to parse Invoice OCR document. To use it, simply upload an image or use the example image below. Results will show up in a few seconds."
article="<b>References</b><br>[1] Y. Xu et al., “LayoutLMv3: Pre-training for Document AI with Unified Text and Image Masking.” 2022. <a href='https://arxiv.org/abs/2204.08387'>Paper Link</a><br>[2] <a href='https://github.com/NielsRogge/Transformers-Tutorials/tree/master/LayoutLMv3'>LayoutLMv3 training and inference</a><br>[3] Hongbin Sun, Zhanghui Kuang, Xiaoyu Yue, Chenhao Lin, and Wayne Zhang. 2021. Spatial Dual-Modality Graph Reasoning for Key Information Extraction. arXiv. DOI:https://doi.org/10.48550/ARXIV.2103.14470 <a href='https://doi.org/10.48550/ARXIV.2103.14470'>Paper Link</a>"
examples =[['example1.png'],['example2.png'],['example3.png'],['inv2.jpg']]
css = """.output_image, .input_image {height: 600px !important}"""
iface = gr.Interface(fn=process_image,
inputs=gr.inputs.Image(type="pil"),
outputs=gr.outputs.Image(type="pil", label="annotated image"),
title=title,
description=description,
article=article,
examples=examples,
css=css,
analytics_enabled = True, enable_queue=True)
iface.launch(inline=False, share=False, debug=True) |