Document Image Classification2020en

rvl-cdip

Dataset from Papers With Code

Metrics:accuracy, cer, wer, f1

Current State of the Art

EAML

Unknown

97.7

accuracy

accuracy Progress Over Time

Showing 5 breakthroughs from Apr 2017 to May 2023

Key Milestones

Apr 2017

Transfer Learning from AlexNet, VGG-16, GoogLeNet and ResNet50

From paper: Cutting the Error by Half: Investigation of Very Deep CNN and Advanced Training Strategies for Document Image Classification

91.0

Source

Jan 2018

Transfer Learning from VGG16 trained on Imagenet

From paper: Document Image Classification with Intra-Domain Transfer Learning and Stacked Generalization of Deep Convolutional Neural Networks

92.2

+1.4%

Source

Dec 2019

Pre-trained LayoutLM

From paper: LayoutLM: Pre-training of Text and Layout for Document Image Understanding

94.4

+2.4%

Source

Jun 2020

Cross-Modal

From paper: Visual and Textual Deep Feature Fusion for Document Image Classification

97.0

+2.8%

Source

May 2023

EAMLCurrent SOTA

From paper: EAML: Ensemble Self-Attention-based Mutual Learning Network for Document Image Classification

97.7

+0.7%

Source

Total Improvement

7.4%

Time Span

6y 2m

Breakthroughs

Current SOTA

97.7

Top Models Performance Comparison

Top 10 models ranked by accuracy

Best Score

97.7

Top Model

EAML

Models Compared

Score Range

2.4

accuracyPrimary

#	Model	Score	Paper / Code	Date
1	EAML	97.7	EAML: Ensemble Self-Attention-based Mutual Learning Network for Document Image Classification	May 2023
2	Cross-Modal	97.05	Visual and Textual Deep Feature Fusion for Document Image Classification	Jun 2020
3	DocFormerBASE	96.17	DocFormer: End-to-End Transformer for Document Understanding Code	Jun 2021
4	LayoutLMV3Large	95.93	LayoutLMv3: Pre-training for Document AI with Unified Text and Image Masking Code	Apr 2022
5	LiLT[EN-R]BASE	95.68	LiLT: A Simple yet Effective Language-Independent Layout Transformer for Structured Document Understanding Code	Feb 2022
6	LayoutLMv2LARGE	95.64	LayoutLMv2: Multi-modal Pre-training for Visually-Rich Document Understanding Code	Dec 2020
7	TILT-Large	95.52	Going Full-TILT Boogie on Document Understanding with Text-Image-Layout Transformer Code	Feb 2021
8	DocFormer large	95.5	DocFormer: End-to-End Transformer for Document Understanding Code	Jun 2021
9	LayoutLMv3BASE	95.44	LayoutLMv3: Pre-training for Document AI with Unified Text and Image Masking Code	Apr 2022
10	Donut	95.3	OCR-free Document Understanding Transformer Code	Nov 2021
11	TILT-Base	95.25	Going Full-TILT Boogie on Document Understanding with Text-Image-Layout Transformer Code	Feb 2021
12	LayoutLMv2BASE	95.25	LayoutLMv2: Multi-modal Pre-training for Visually-Rich Document Understanding Code	Dec 2020
13	LayoutXLM	95.21	LayoutXLM: Multimodal Pre-training for Multilingual Visually-rich Document Understanding Code	Apr 2021
14	StrucTexTv2 (large)	94.62	StrucTexTv2: Masked Visual-Textual Prediction for Document Image Pre-training Code	Mar 2023
15	Pre-trained LayoutLM	94.42	LayoutLM: Pre-training of Text and Layout for Document Image Understanding Code	Dec 2019
16	DoPTA	94.12	DoPTA: Improving Document Layout Analysis using Patch-Text Alignment	Dec 2024
17	DocXClassifier-B	94	DocXClassifier: High Performance Explainable Deep Network for Document Image ClassificationCode	Mar 2022
18	StrucTexTv2 (small)	93.4	StrucTexTv2: Masked Visual-Textual Prediction for Document Image Pre-training Code	Mar 2023
19	VLCDoC	93.19	VLCDoC: Vision-Language Contrastive Pre-Training Model for Cross-Modal Document Classification	May 2022
20	TransferDoc	93.18	GlobalDoc: A Cross-Modal Vision-Language Framework for Real-World Document Image Retrieval and Classification	Sep 2023
21	Multimodal (ResNet50)	92.7	Multimodal Side-Tuning for Document Classification Code	Jan 2023
22	DiT-L	92.69	DiT: Self-supervised Pre-training for Document Image Transformer Code	Mar 2022
23	Pre-trained EfficientNet	92.31	Improving accuracy and speeding up Document Image Classification through parallel systems Code	Jun 2020
24	Transfer Learning from VGG16 trained on Imagenet	92.21	Document Image Classification with Intra-Domain Transfer Learning and Stacked Generalization of Deep Convolutional Neural Networks Code	Jan 2018
25	Multimodal (MobileNetV2)	92.2	Multimodal Side-Tuning for Document Classification Code	Jan 2023
26	DiT-B	92.11	DiT: Self-supervised Pre-training for Document Image Transformer Code	Mar 2022
27	BEiT-B	91.09	BEiT: BERT Pre-Training of Image Transformers Code	Jun 2021
28	Transfer Learning from AlexNet, VGG-16, GoogLeNet and ResNet50	90.97	Cutting the Error by Half: Investigation of Very Deep CNN and Advanced Training Strategies for Document Image Classification Code	Apr 2017
29	AlexNet + spatial pyramidal pooling + image resizing	90.94	Analysis of Convolutional Neural Networks for Document Image Classification	Aug 2017
30	DeiT-BOpen Source Meta	90.32	Training data-efficient image transformers & distillation through attention Code	Dec 2020
31	Roberta base	90.06	RoBERTa: A Robustly Optimized BERT Pretraining Approach Code	Jul 2019