cvdataset-layoutlmv3.py

import json
import os
from pathlib import Path
import datasets
from PIL import Image
import pandas as pd

logger = datasets.logging.get_logger(__name__)
_CITATION = """\
@article{LayoutLmv3 for CV extractions,
  title={LayoutLmv3for Key Information Extraction},
  author={MisaR&D Team},
  year={2022},
}
"""
_DESCRIPTION = """\
CV is a collection of receipts. It contains, for each photo about cv personal, a list of OCRs - with the bounding box, text, and class. The goal is to benchmark "key information extraction" - extracting key information from documents
https://arxiv.org/abs/2103.14470
"""

def load_image(image_path):
    image = Image.open(image_path).convert("RGB")
    w, h = image.size
    return image, (w,h)

def normalize_bbox(bbox, size):
    return [
        int(1000 * bbox[0] / size[0]),
        int(1000 * bbox[1] / size[1]),
        int(1000 * bbox[2] / size[0]),
        int(1000 * bbox[3] / size[1]),
    ]

def _get_drive_url(url):
    base_url = 'https://drive.google.com/uc?id='
    split_url = url.split('/')
    
    return base_url + split_url[5]
    
_URLS = [
    _get_drive_url("https://drive.google.com/file/d/11SRDeRKUr8XacB7tauiGjkw1PXDGFKUx/"),
    _get_drive_url("https://drive.google.com/file/d/1KdDBmGP96lFc7jv2Bf4eqrO121ST-TCh/"),
]

class DatasetConfig(datasets.BuilderConfig):
    """BuilderConfig for WildReceipt Dataset"""
    def __init__(self, **kwargs):
        """BuilderConfig for WildReceipt Dataset.
        Args:
          **kwargs: keyword arguments forwarded to super.
        """
        super(DatasetConfig, self).__init__(**kwargs)


class WildReceipt(datasets.GeneratorBasedBuilder):
    BUILDER_CONFIGS = [
        DatasetConfig(name="CV Extractions", version=datasets.Version("1.0.0"), description="CV dataset"),
    ]

    def _info(self):
        return datasets.DatasetInfo(
            description=_DESCRIPTION,
            features=datasets.Features(
                {
                    "id": datasets.Value("string"),
                    "words": datasets.Sequence(datasets.Value("string")),
                    "bboxes": datasets.Sequence(datasets.Sequence(datasets.Value("int64"))),
                    "ner_tags": datasets.Sequence(
                        datasets.features.ClassLabel(
                            names=['person_name', 'dob_key', 'dob_value', 'gender_key', 'gender_value', 'phonenumber_key', 'phonenumber_value', 'email_key', 'email_value', 'address_key', 'address_value', 'socical_address_value', 'education', 'education_name', 'education_time', 'experience', 'experience_name', 'experience_time', 'information', 'undefined']
                        )
                    ),
                    "image_path": datasets.Value("string"),
                }
            ),
            supervised_keys=None,
            citation=_CITATION,
            homepage="",
        )

    def _split_generators(self, dl_manager):
        """Returns SplitGenerators."""
        """Uses local files located with data_dir"""
        downloaded_file = dl_manager.download_and_extract(_URLS)
        dest = Path(downloaded_file[0])/'data1'

        return [
            datasets.SplitGenerator(
                name=datasets.Split.TRAIN, gen_kwargs={"filepath": dest/"train.txt", "dest": dest}
            ),            
            datasets.SplitGenerator(
                name=datasets.Split.TEST, gen_kwargs={"filepath": dest/"test.txt", "dest": dest}
            ),
        ]

    def _generate_examples(self, filepath, dest):

        df = pd.read_csv(dest/'class_list.txt', delimiter='\s', header=None)
        id2labels = dict(zip(df[0].tolist(), df[1].tolist()))


        logger.info("⏳ Generating examples from = %s", filepath)

        item_list = []
        with open(filepath, 'r') as f:
            for line in f:
                item_list.append(line.rstrip('\n\r'))
        
        for guid, fname in enumerate(item_list):

            data = json.loads(fname)
            image_path = dest/data['file_name']
            image, size = load_image(image_path)
            boxes = [[i['box'][6], i['box'][7], i['box'][2], i['box'][3]] for i in data['annotations']]

            text = [i['text'] for i in data['annotations']]
            label = [id2labels[i['label']] for i in data['annotations']]
            
            boxes = [normalize_bbox(box, size) for box in boxes]
            
            flag=0
            for i in boxes:
              for j in i:
                if j>1000:
                  flag+=1
                  pass
            if flag>0: print(image_path)
 
            yield guid, {"id": str(guid), "words": text, "bboxes": boxes, "ner_tags": label, "image_path": image_path}