LayoutLM¶
Overview¶
The LayoutLM model was proposed in the paper LayoutLM: Pre-training of Text and Layout for Document Image Understanding by Yiheng Xu, Minghao Li, Lei Cui, Shaohan Huang, Furu Wei, and Ming Zhou. It’s a simple but effective pretraining method of text and layout for document image understanding and information extraction tasks, such as form understanding and receipt understanding. It obtains state-of-the-art results on several downstream tasks:
form understanding: the FUNSD dataset (a collection of 199 annotated forms comprising more than 30,000 words).
receipt understanding: the SROIE dataset (a collection of 626 receipts for training and 347 receipts for testing).
document image classification: the RVL-CDIP dataset (a collection of 400,000 images belonging to one of 16 classes).
The abstract from the paper is the following:
Pre-training techniques have been verified successfully in a variety of NLP tasks in recent years. Despite the widespread use of pretraining models for NLP applications, they almost exclusively focus on text-level manipulation, while neglecting layout and style information that is vital for document image understanding. In this paper, we propose the LayoutLM to jointly model interactions between text and layout information across scanned document images, which is beneficial for a great number of real-world document image understanding tasks such as information extraction from scanned documents. Furthermore, we also leverage image features to incorporate words’ visual information into LayoutLM. To the best of our knowledge, this is the first time that text and layout are jointly learned in a single framework for document-level pretraining. It achieves new state-of-the-art results in several downstream tasks, including form understanding (from 70.72 to 79.27), receipt understanding (from 94.02 to 95.24) and document image classification (from 93.07 to 94.42).
Tips:
In addition to input_ids,
forward()also expects the inputbbox, which are the bounding boxes (i.e. 2D-positions) of the input tokens. These can be obtained using an external OCR engine such as Google’s Tesseract (there’s a Python wrapper available). Each bounding box should be in (x0, y0, x1, y1) format, where (x0, y0) corresponds to the position of the upper left corner in the bounding box, and (x1, y1) represents the position of the lower right corner. Note that one first needs to normalize the bounding boxes to be on a 0-1000 scale. To normalize, you can use the following function:
def normalize_bbox(bbox, width, height):
return [
int(1000 * (bbox[0] / width)),
int(1000 * (bbox[1] / height)),
int(1000 * (bbox[2] / width)),
int(1000 * (bbox[3] / height)),
]
Here, width and height correspond to the width and height of the original document in which the token
occurs. Those can be obtained using the Python Image Library (PIL) library for example, as follows:
from PIL import Image
image = Image.open("name_of_your_document - can be a png file, pdf, etc.")
width, height = image.size
For a demo which shows how to fine-tune
LayoutLMForTokenClassificationon the FUNSD dataset (a collection of annotated forms), see this notebook. It includes an inference part, which shows how to use Google’s Tesseract on a new document.
The original code can be found here.