Optical Character Recognition2020en

imdb-m

Dataset from Papers With Code

Metrics:accuracy, cer, wer, f1
Current State of the Art

Document Classification Using Importance of Sentences

Unknown

54.8

accuracy

accuracy Progress Over Time

Showing 2 breakthroughs from Jun 2019 to Mar 2021

52.653.253.854.455.0Jun 2019Mar 2021accuracyDate

Key Milestones

Jun 2019
LSTM-reg (single model)

From paper: Rethinking Complex Neural Network Architectures for Document Classification

52.8
Mar 2021
Document Classification Using Importance of SentencesCurrent SOTA

From paper: Improving Document-Level Sentiment Classification Using Importance of Sentences

54.8
+3.8%
Total Improvement
3.8%
Time Span
1y 9m
Breakthroughs
2
Current SOTA
54.8

accuracyPrimary

#ModelScorePaper / CodeDate
1
Document Classification Using Importance of Sentences
54.8Mar 2021
2
LSTM-reg (single model)
52.8
Rethinking Complex Neural Network Architectures for Document ClassificationCode
Jun 2019

Related Papers1

Improving Document-Level Sentiment Classification Using Importance of Sentences
Mar 2021Models: Document Classification Using Importance of Sentences

Other Optical Character Recognition Datasets

imdb-m Benchmark - Optical Character Recognition | CodeSOTA