Scene Text Recognition On Icdar2015

Metrics

Accuracy

Results

Performance results of various models on this benchmark

Model Name	Accuracy	Paper Title	Repository
PARSeq	89.6±0.3	Scene Text Recognition with Permuted Autoregressive Sequence Models	-
DAN	74.5	Decoupled Attention Network for Text Recognition	-
SAR	69.2	Show, Attend and Read: A Simple and Strong Baseline for Irregular Text Recognition	-
AON	73.0	AON: Towards Arbitrarily-Oriented Text Recognition	-
ASTER	76.1	ASTER: An Attentional Scene Text Recognizer with Flexible Rectification
DTrOCR 105M	93.5	DTrOCR: Decoder-only Transformer for Optical Character Recognition	-
SIGA_S	87.6	Self-supervised Implicit Glyph Attention for Text Recognition	-
TextScanner	79.4	TextScanner: Reading Characters in Order for Robust Scene Text Recognition	-
CLIP4STR-L (DataComp-1B)	91.4	CLIP4STR: A Simple Baseline for Scene Text Recognition with Pre-trained Vision-Language Model	-
Baek et al.	71.8	What Is Wrong With Scene Text Recognition Model Comparisons? Dataset and Model Analysis	-
CLIP4STR-L	90.8	CLIP4STR: A Simple Baseline for Scene Text Recognition with Pre-trained Vision-Language Model	-
DPAN	85.5	Look Back Again: Dual Parallel Attention Network for Accurate and Robust Scene Text Recognition
S-GTR	87.3	Visual Semantics Allow for Textual Reasoning Better in Scene Text Recognition	-
CLIP4STR-B	90.6	CLIP4STR: A Simple Baseline for Scene Text Recognition with Pre-trained Vision-Language Model	-
CLIP4STR-L*	92.6	An Empirical Study of Scaling Law for OCR	-
MGP-STR	90.9	Multi-Granularity Prediction for Scene Text Recognition	-
ViTSTR	72.6	Vision Transformer for Fast and Efficient Scene Text Recognition	-
CSTR	81.6	Revisiting Classification Perspective on Scene Text Recognition	-
SAFL	77.5	SAFL: A Self-Attention Scene Text Recognizer with Focal Loss	-
Yet Another Text Recognizer	80.2	Why You Should Try the Real Data for the Scene Text Recognition	-

0 of 27 row(s) selected.