Scene Text Recognition2020en

iiit5k

Dataset from Papers With Code

Metrics:accuracy, cer, wer, f1

Current State of the Art

CLIP4STR-L (DataComp-1B)

Unknown

99.6

accuracy

accuracy Progress Over Time

Showing 6 breakthroughs from Jul 2015 to May 2023

Key Milestones

Jul 2015

CRNN

Lexicon-free. Table 2. TPAMI 2017.

78.2

Source

Mar 2021

ABINet-LV

ABINet Language-Vision variant. CVPR 2021.

96.4

+23.3%

Source

Nov 2021

MATRN

From paper: Multi-modal Text Recognition Networks: Interactive Enhancements between Visual and Semantic Features

96.6

+0.2%

Source

Dec 2021

S-GTR

From paper: Visual Semantics Allow for Textual Reasoning Better in Scene Text Recognition

97.5

+0.9%

Source

Jul 2022

PARSeq

Lowercase alphanum eval, 3000 test samples. ECCV 2022.

99.0

+1.5%

Source

May 2023

CLIP4STR-L (DataComp-1B)Current SOTA

From paper: CLIP4STR: A Simple Baseline for Scene Text Recognition with Pre-trained Vision-Language Model

99.6

+0.6%

Source

Total Improvement

27.4%

Time Span

7y 11m

Breakthroughs

Current SOTA

99.6

Top Models Performance Comparison

Top 10 models ranked by accuracy

Best Score

99.6

Top Model

CLIP4STR-L (DataC...

Models Compared

Score Range

1.6

accuracyPrimary

#	Model	Score	Paper / Code	Date
1	CLIP4STR-L (DataComp-1B)	99.6	CLIP4STR: A Simple Baseline for Scene Text Recognition with Pre-trained Vision-Language Model Code	May 2023
2	DTrOCR 105M	99.6	DTrOCR: Decoder-only Transformer for Optical Character Recognition Code	Aug 2023
3	CLIP4STR-L	99.5	CLIP4STR: A Simple Baseline for Scene Text Recognition with Pre-trained Vision-Language Model Code	May 2023
4	CLIP4STR-B (DataComp-1B)	99.5	CLIP4STR: A Simple Baseline for Scene Text Recognition with Pre-trained Vision-Language Model Code	May 2023
5	CPPD	99.3	Context Perception Parallel Decoder for Scene Text Recognition Code	Jul 2023
6	CLIP4STR-B Research	99.2	CLIP4STR: A Simple Baseline for Scene Text Recognition with Pre-trained Vision-Language Model Code	May 2023
7	PARSeqOpen Source Research	99	Scene Text Recognition with Permuted Autoregressive Sequence Models	Jul 2022
8	MGP-STR	98.8	Multi-Granularity Prediction for Scene Text Recognition Code	Sep 2022
9	CCD-ViT-Small(ARD_2.8M)	98	Self-supervised Character-to-Character Distillation for Text Recognition Code	Nov 2022
10	CCD-ViT-Base(ARD_2.8M)	98	Self-supervised Character-to-Character Distillation for Text Recognition Code	Nov 2022
11	S-GTR	97.5	Visual Semantics Allow for Textual Reasoning Better in Scene Text Recognition Code	Dec 2021
12	DiffusionSTR	97.3	DiffusionSTR: Diffusion Model for Scene Text Recognition	Jun 2023
13	CCD-ViT-Tiny(ARD_2.8M)	97.1	Self-supervised Character-to-Character Distillation for Text Recognition Code	Nov 2022
14	SIGA_S	96.9	Self-supervised Implicit Glyph Attention for Text Recognition Code	Mar 2022
15	MATRN Research	96.6	Multi-modal Text Recognition Networks: Interactive Enhancements between Visual and Semantic Features Code	Nov 2021
16	CDistNet (Ours)	96.57	CDistNet: Perceiving Multi-Domain Character Distance for Robust Text Recognition Code	Nov 2021
17	ABINet-LVOpen Source Fang et al.	96.4	Read Like Humans: Autonomous, Bidirectional and Iterative Language Modeling for Scene Text Recognition	Mar 2021
18	DPAN	96.2	Look Back Again: Dual Parallel Attention Network for Accurate and Robust Scene Text RecognitionCode	Aug 2021
19	TrOCR-large 558M	94.1	TrOCR: Transformer-based Optical Character Recognition with Pre-trained Models	Sep 2021
20	TrOCR-base 334M	93.4	TrOCR: Transformer-based Optical Character Recognition with Pre-trained Models	Sep 2021
21	CRNN	78.2	An End-to-End Trainable Neural Network for Image-based Sequence Recognition and Its Application to Scene Text Recognition	Jul 2015