Skip to content

Latest commit

 

History

History
36 lines (29 loc) · 717 Bytes

README.md

File metadata and controls

36 lines (29 loc) · 717 Bytes

InvoiceNet

Classification of entities extracted from invoice scans

Instructions

Two models have been implemented for the purpose of invoice data classification.

The first one is a convolution neural network which trains on custom word embeddings trained using word2vec.

python3 main.py -h

The second model is based on CloudScan.

python3 cloud_scan.py -h

Dependencies:

python3

nltk
pandas
numpy
keras
pickle
gzip
gensim
sklearn
matplotlib
tqdm
datefinder

References:

  • Rasmus Berg Palm, Ole Winther, Florian Laws, CloudScan - A configuration-free invoice analysis system using recurrent neural networks, arXiv:1708.07403v1 [cs.CL]