Text Spotting on Total-Text

13개 결과 · ⬇ CSV · JSON

F-measure (%) - No Lexicon

Rank	Model	F-measure (%) - No Lexicon	F-measure (%) - Full Lexicon	Extra Training Data	Paper	Code	Year
1	DeepSolo (ViTAEv2-S, TextOCR)	83.6	89.6	✓	DeepSolo: Let Transformer Decoder with Explicit Points Solo for Text Spotting	vitae-transformer/deepsolo	2022
2	DeepSolo (ResNet-50, TextOCR)	82.5	88.7	✓	DeepSolo: Let Transformer Decoder with Explicit Points Solo for Text Spotting	vitae-transformer/deepsolo	2022
3	DeepSolo (ResNet-50)	79.7	87.0		DeepSolo: Let Transformer Decoder with Explicit Points Solo for Text Spotting	vitae-transformer/deepsolo	2022
4	A3S	79.4	85.1		A3S: Adversarial learning of semantic representations for Scene-Text Spotting		2023
5	UNITS	78.7	86.0	✓	Towards Unified Scene Text Spotting based on Sequence Generation	clovaai/units	2023
6	GLASS	76.6	83.0		GLASS: Global to Local Attention for Scene-Text Spotting	amazon-science/glass-text-spotting · amazon-research/glass-text-spotting	2022
7	DEER	74.8	83.3		DEER: Detection-agnostic End-to-End Recognizer for Scene Text Spotting		2022
8	SwinTextSpotter	74.3	84.1		SwinTextSpotter: Scene Text Spotting via Better Synergy between Text Detection and Text Recognition	mxin262/swintextspotter · jacobtyo/swintextspotter	2022
9	TESTR	73.3	83.9		Text Spotting Transformers	mlpc-ucsd/testr	2022
10	MANGO	72.9	83.6		MANGO: A Mask Attention Guided One-Stage Scene Text Spotter	hikopensource/davar-lab-ocr	2020
11	MaskTextSpotter v3	71.2	78.4		Mask TextSpotter v3: Segmentation Proposal Network for Robust Scene Text Spotting	MhLiao/MaskTextSpotterV3	2020
12	ABCNet v2	70.4	78.1		ABCNet v2: Adaptive Bezier-Curve Network for Real-time End-to-end Text Spotting	Yuliang-Liu/ABCNet_Chinese	2021
13	Soft 자동 추출	–	83.4		Do You Need Text Rectification? Soft Attention Mask Embedding for Rectification-Free Scene Text Spotting		2026