SOTAVerified

Scene Text Recognition

See Scene Text Detection for leaderboards in this task.

Papers

Recently AddedMost HypedMost ActiveNeeds VerificationMost Verified

Showing 101–150 of 269 papers

TitleStatusHype
SGBANet: Semantic GAN and Balanced Attention Network for Arbitrarily Oriented Scene Text Recognition—0
Scene Text Recognition with Permuted Autoregressive Sequence ModelsCode2
Reading and Writing: Discriminative and Generative Modeling for Self-Supervised Text RecognitionCode1
GIT: A Generative Image-to-text Transformer for Vision and LanguageCode2
Multimodal Semi-Supervised Learning for Text RecognitionCode1
SVTR: Scene Text Recognition with a Single Visual ModelCode0
Pushing the Performance Limit of Scene Text Recognizer without Human AnnotationCode1
IterVM: Iterative Vision Modeling Module for Scene Text RecognitionCode1
SimAN: Exploring Self-Supervised Representation Learning of Scene Text via Similarity-Aware NormalizationCode1
Training Protocol Matters: Towards Accurate Scene Text Recognition via Training Protocol SearchingCode1
Towards Open-Set Text Recognition via Label-to-Prototype Learning—0
Text-DIAE: A Self-Supervised Degradation Invariant Autoencoders for Text Recognition and Document EnhancementCode1
Self-supervised Implicit Glyph Attention for Text RecognitionCode1
Towards Boosting the Accuracy of Non-Latin Scene Text RecognitionCode0
Transfer Learning for Scene Text Recognition in Indian Languages—0
On the Cross-dataset Generalization in License Plate RecognitionCode1
SAFL: A Self-Attention Scene Text Recognizer with Focal LossCode0
Visual Semantics Allow for Textual Reasoning Better in Scene Text RecognitionCode1
Text Gestalt: Stroke-Aware Scene Text Image Super-ResolutionCode1
Visual-Semantic Transformer for Scene Text Recognition—0
Multi-modal Text Recognition Networks: Interactive Enhancements between Visual and Semantic FeaturesCode1
Traditional Chinese Synthetic Datasets Verified with Labeled Data for Scene Text RecognitionCode1
Decoupling Visual-Semantic Feature Learning for Robust Scene Text Recognition—0
Utilizing Resource-Rich Language Datasets for End-to-End Scene Text Recognition in Resource-Poor Languages—0
CDistNet: Perceiving Multi-Domain Character Distance for Robust Text RecognitionCode1
TRIG: Transformer-Based Text Recognizer with Initial Embedding Guidance—0
Oracle Teacher: Leveraging Target Information for Better Knowledge Distillation of CTC Models—0
TPSNet: Reverse Thinking of Thin Plate Splines for Arbitrary Shape Scene Text RepresentationCode1
Ultra Light OCR Competition Technical Report—0
Scene Text Image Super-Resolution via Parallelly Contextual Attention NetworkCode0
Sharp Attention for Sequence to Sequence Learning—0
TrOCR: Transformer-based Optical Character Recognition with Pre-trained ModelsCode1
PIMNet: A Parallel, Iterative and Mimicking Network for Scene Text RecognitionCode1
Meta Self-Learning for Multi-Source Domain Adaptation: A BenchmarkCode1
From Two to One: A New Scene Text Recognizer with Visual Language Modeling NetworkCode1
Data Augmentation for Scene Text RecognitionCode1
IFR: Iterative Fusion Based Recognizer For Low Quality Scene Text Recognition—0
Look Back Again: Dual Parallel Attention Network for Accurate and Robust Scene Text RecognitionCode0
Why You Should Try the Real Data for the Scene Text RecognitionCode0
Arabic Scene Text Recognition in the Deep Learning Era: Analysis on A Novel DatasetCode1
Text is Text, No Matter What: Unifying Text Recognition using Knowledge Distillation—0
RewriteNet: Reliable Scene Text Editing with Implicit Decomposition of Text Contents and StylesCode0
SynthTIGER: Synthetic Text Image GEneratoR Towards Better Text Recognition ModelsCode1
Scene Text recognition with Full Normalization—0
CentripetalText: An Efficient Text Instance Representation for Scene Text DetectionCode1
Scene Text Telescope: Text-Focused Scene Image Super-ResolutionCode0
Dictionary-Guided Scene Text RecognitionCode1
Representation and Correlation Enhanced Encoder-Decoder Framework for Scene Text RecognitionCode0
Convolutional Neural Networks with Gated Recurrent ConnectionsCode1
Vision Transformer for Fast and Efficient Scene Text RecognitionCode1
Show:102550
← PrevPage 3 of 6Next →
All datasetsICDAR2013SVTICDAR2015CUTE80IIIT5kSVTPICDAR 2003WOSTIC19-ArtCOCO-TextHOSTUber-Text

Benchmark Results

#ModelMetricClaimedVerifiedStatus
1CLIP4STR-L*Accuracy99.42—Unverified
2DTrOCR 105MAccuracy99.4—Unverified
3CLIP4STR-L (DataComp-1B)Accuracy99—Unverified
4CLIP4STR-LAccuracy98.5—Unverified
5MGP-STRAccuracy98.5—Unverified
6CCD-ViT-Small(ARD_2.8M)Accuracy98.3—Unverified
7CLIP4STR-BAccuracy98.3—Unverified
8CCD-ViT-Base(ARD_2.8M)Accuracy98.3—Unverified
9MATRNAccuracy97.9—Unverified
10S-GTRAccuracy97.8—Unverified
#ModelMetricClaimedVerifiedStatus
1CLIP4STR-H (DFN-5B)Accuracy99.1—Unverified
2DTrOCR 105MAccuracy98.9—Unverified
3CLIP4STR-B*Accuracy98.76—Unverified
4CLIP4STR-L (DataComp-1B)Accuracy98.6—Unverified
5MGP-STRAccuracy98.6—Unverified
6CLIP4STR-LAccuracy98.5—Unverified
7CPPDAccuracy98.5—Unverified
8CLIP4STR-BAccuracy98.3—Unverified
9CCD-ViT-Base(ARD_2.8M)Accuracy97.8—Unverified
10CCD-ViT-Small(ARD_2.8M)Accuracy96.4—Unverified
#ModelMetricClaimedVerifiedStatus
1DTrOCR 105MAccuracy93.5—Unverified
2CLIP4STR-L*Accuracy92.6—Unverified
3CPPDAccuracy91.7—Unverified
4CLIP4STR-L (DataComp-1B)Accuracy91.4—Unverified
5MGP-STRAccuracy90.9—Unverified
6CLIP4STR-LAccuracy90.8—Unverified
7CLIP4STR-BAccuracy90.6—Unverified
8SIGA_SAccuracy87.6—Unverified
9S-GTRAccuracy87.3—Unverified
10MATRNAccuracy86.6—Unverified
#ModelMetricClaimedVerifiedStatus
1CPPDAccuracy99.7—Unverified
2CLIP4STR-L (DataComp-1B)Accuracy99.7—Unverified
3CLIP4STR-B*Accuracy99.65—Unverified
4MGP-STRAccuracy99.31—Unverified
5CLIP4STR-BAccuracy99.3—Unverified
6DTrOCR 105MAccuracy99.1—Unverified
7CLIP4STR-LAccuracy99—Unverified
8CCD-ViT-Base(ARD_2.8M)Accuracy98.3—Unverified
9CCD-ViT-Small(ARD_2.8M)Accuracy98.3—Unverified
10CCD-ViT-Tiny(ARD_2.8M)Accuracy95.8—Unverified
#ModelMetricClaimedVerifiedStatus
1CLIP4STR-L (DataComp-1B)Accuracy99.6—Unverified
2DTrOCR 105MAccuracy99.6—Unverified
3CLIP4STR-B (DataComp-1B)Accuracy99.5—Unverified
4CLIP4STR-LAccuracy99.5—Unverified
5CPPDAccuracy99.3—Unverified
6CLIP4STR-BAccuracy99.2—Unverified
7MGP-STRAccuracy98.8—Unverified
8CCD-ViT-Base(ARD_2.8M)Accuracy98—Unverified
9CCD-ViT-Small(ARD_2.8M)Accuracy98—Unverified
10S-GTRAccuracy97.5—Unverified
#ModelMetricClaimedVerifiedStatus
1DTrOCR 105MAccuracy98.6—Unverified
2MGP-STRAccuracy98.3—Unverified
3CLIP4STR-L*Accuracy98.13—Unverified
4CLIP4STR-L (DataComp-1B)Accuracy98.1—Unverified
5CLIP4STR-LAccuracy97.4—Unverified
6CLIP4STR-BAccuracy97.2—Unverified
7CPPDAccuracy96.7—Unverified
8CCD-ViT-BaseAccuracy96.1—Unverified
9CCD-ViT-SmallAccuracy92.7—Unverified
10CCD-ViT-TinyAccuracy91.6—Unverified
#ModelMetricClaimedVerifiedStatus
1Yet Another Text RecognizerAccuracy97.1—Unverified
2SIGA_TAccuracy97—Unverified
3SATRNAccuracy96.7—Unverified
4DANAccuracy95—Unverified
5SAFLAccuracy95—Unverified
6CSTRAccuracy94.8—Unverified
7Baek et al.Accuracy94.4—Unverified
8ViTSTRAccuracy94.3—Unverified
9AONAccuracy91.5—Unverified
10RAREAccuracy90.1—Unverified
#ModelMetricClaimedVerifiedStatus
1CLIP4STR-H (DFN-5B)1:1 Accuracy90.9—Unverified
2CLIP4STR-L (DataComp-1B)1:1 Accuracy90.6—Unverified
3CLIP4STR-L1:1 Accuracy88.8—Unverified
4CLIP4STR-B1:1 Accuracy87—Unverified
5CCD-ViT-Base1:1 Accuracy86—Unverified
#ModelMetricClaimedVerifiedStatus
1CLIP4STR-L (DataComp-1B)Accuracy (%)86.4—Unverified
2CLIP4STR-LAccuracy (%)85.9—Unverified
3CLIP4STR-BAccuracy (%)85.8—Unverified
4MGP-STRAccuracy (%)85.5—Unverified
#ModelMetricClaimedVerifiedStatus
1CLIP4STR-L1:1 Accuracy81.9—Unverified
2MGP-STR1:1 Accuracy81.7—Unverified
3CLIP4STR-B1:1 Accuracy81.1—Unverified
#ModelMetricClaimedVerifiedStatus
1CLIP4STR-L1:1 Accuracy82.7—Unverified
2CLIP4STR-B1:1 Accuracy79.8—Unverified
3CCD-ViT-Base1:1 Accuracy77.3—Unverified
#ModelMetricClaimedVerifiedStatus
1CLIP4STR-L (DataComp-1B)Accuracy (%)92.2—Unverified
2MGP-STRAccuracy (%)91—Unverified
3CLIP4STR-BAccuracy (%)86.8—Unverified
#ModelMetricClaimedVerifiedStatus
1ABINet-LV+TPS++Accuracy97.8—Unverified
#ModelMetricClaimedVerifiedStatus
1MLDGAverage Accuracy19.02—Unverified
#ModelMetricClaimedVerifiedStatus
1ABINet-LV+TPS++Accuracy89.6—Unverified