Accueil Actualités Articles de recherche récents Tutoriels Ensembles de données Wiki SOTA Modèles LLM Classement GPU Événements

Français

Scene Text Detection On Msra Td500

Métriques

F-Measure

FPS

Precision

Recall

Résultats

Résultats de performance de divers modèles sur ce benchmark

Nom du modèle	F-Measure	FPS	Precision	Recall	Paper Title	Repository
MixNet	89.4	15.2	90.7	88.1	MixNet: Toward Accurate Detection of Challenging Scene Text in the Wild
DB-ResNet-50 (736)	84.9	-	91.5	79.2	Real-time Scene Text Detection with Differentiable Binarization
TextSnake	78.3	-	83.2	73.9	TextSnake: A Flexible Representation for Detecting Text of Arbitrary Shapes
EAST + PVANET2x	76.08	-	87.28	67.43	EAST: An Efficient and Accurate Scene Text Detector
FAST-T-512	84.5	137.2	91.1	78.8	FAST: Faster Arbitrarily-Shaped Text Detector with Minimalist Kernel Representation
PAN	84.1	-	-	83.8	Efficient and Accurate Arbitrary-Shaped Text Detection with Pixel Aggregation Network
FAST-B-736	87.3	56.8	92.1	83	FAST: Faster Arbitrarily-Shaped Text Detector with Minimalist Kernel Representation
RRD∗	79	-	87	73	Rotation-Sensitive Regression for Oriented Scene Text Detection	-
PixelLink + VGG16 2s	77.8	-	83	73.2	PixelLink: Detecting Scene Text via Instance Segmentation
SegLink	77	-	86	70	Detecting Oriented Text in Natural Images by Linking Segments
Corner Localization	81.5	-	87.6	76.2	Multi-Oriented Scene Text Detection via Corner Localization and Region Segmentation
DBNet++ (ResNet-18) (512)	82.6	80	89.7	76.5	Real-Time Scene Text Detection with Differentiable Binarization and Adaptive Scale Fusion
DBNet++ (ResNet-50) (736)	87.2	29	91.5	83.3	Real-Time Scene Text Detection with Differentiable Binarization and Adaptive Scale Fusion
DBNet++ (ResNet-18) (736)	85.1	55	87.9	82.5	Real-Time Scene Text Detection with Differentiable Binarization and Adaptive Scale Fusion
FAST-T-736	84.9	79.6	88.1	81.9	FAST: Faster Arbitrarily-Shaped Text Detector with Minimalist Kernel Representation
FAST-S-736	86.4	72	91.6	81.7	FAST: Faster Arbitrarily-Shaped Text Detector with Minimalist Kernel Representation
CRAFT	82.9	-	88.2	78.2	Character Region Awareness for Text Detection
FTSN + MNMS	82	-	87.6	77.1	Fused Text Segmentation Networks for Multi-oriented Scene Text Detection	-

0 of 18 row(s) selected.