adlnlp/doc_gcn
github.com · 581 words · saved by 1 readers
Contribute to adlnlp/doc_gcn development by creating an account on GitHub.
Doc-GCN: Heterogeneous Graph Convolutional Networks for Document Layout Analysis This repository contains code for the paper Doc-GCN: Heterogeneous Graph Convolutional Networks for Document Layout Analysis. Siwen Luo*, Yihao Ding*, Siqu Long, Soyeon Caren Han, Josiah Poon University of Sydney and University of Western Australia Dataset Prepare This paper uses three widely used benchmark datasets, including FUNSD (paper), Publaynet (paper), and Docbank (paper). (All three datasets are publicly available and can be gotten via their officially provided download link.) Before feeding into…
saved by
related reading
- A Gentle Introduction to Graph Neural Networksdistill.pub
- Understanding Convolutions on Graphsdistill.pub
- Graph Convolutional Networks | Thomas Kipf | Google DeepMindtkipf.github.io
- [2207.02505] Pure Transformers are Powerful Graph Learnersarxiv.org
- Circuit Tracing: Revealing Computational Graphs in Language Modelstransformer-circuits.pub
- Distill — Latest articles about machine learningdistill.pub
- Aman's AI Journal • Primers • Ilya Sutskever's Top 30aman.ai
- title.txtsybrandt.com
- [1801.07606] Deeper Insights into Graph Convolutional Networks for Semi-Supervised Learningarxiv.org
- NLPContributionGraph -- Structuring Scholarly NLP Contributions in the Open Research Knowledge Graphncg-task.github.io
- [1912.13318] LayoutLM: Pre-training of Text and Layout for Document Image Understandingarxiv.org
- Beyond Message Passing: a Physics-Inspired Paradigm for Graph Neural Networksthegradient.pub