SOTAVerified

Handwritten Text Recognition

Handwritten Text Recognition (HTR) is the task of automatically identifying and transcribing handwritten text from images or scanned documents into machine-readable text. The goal is to develop a system capable of accurately interpreting diverse handwriting styles, accounting for variations in alignment, stroke, spacing, and noise. This task involves detecting handwritten regions within an image, extracting the text content, and converting it into a structured digital format, enabling further search, indexing, or data analysis.

Papers

Recently AddedMost HypedMost ActiveNeeds VerificationMost Verified

Showing 1–50 of 139 papers

TitleStatusHype
Full Page Handwriting Recognition via Image to Sequence ExtractionCode2
DTrOCR: Decoder-only Transformer for Optical Character RecognitionCode2
HTR-VT: Handwritten Text Recognition with Vision TransformerCode2
Unlocking the Archives: Using Large Language Models to Transcribe Handwritten Historical DocumentsCode2
StackMix and Blot Augmentations for Handwritten Text RecognitionCode1
Decoupled Attention Network for Text RecognitionCode1
Data Generation for Post-OCR correction of Cyrillic handwritingCode1
ScrabbleGAN: Semi-Supervised Varying Length Handwritten Text GenerationCode1
TrOCR: Transformer-based Optical Character Recognition with Pre-trained ModelsCode1
Towards End-to-end Handwritten Document RecognitionCode1
Few Shots Are All You Need: A Progressive Few Shot Learning Approach for Low Resource Handwritten Text RecognitionCode1
Stylometry for Noisy Medieval Data: Evaluating Paul Meyer's Hagiographic HypothesisCode1
Recurrence-free unconstrained handwritten text recognition using gated fully convolutional networkCode1
Digital Peter: Dataset, Competition and Handwriting Recognition MethodsCode1
Many Heads but One Brain: Fusion Brain -- a Competition and a Single Multimodal Multitask ArchitectureCode1
MetaHTR: Towards Writer-Adaptive Handwritten Text RecognitionCode1
LineCounter: Learning Handwritten Text Line Segmentation by CountingCode1
Meta-DAN: towards an efficient prediction strategy for page-level handwritten text recognitionCode1
OrigamiNet: Weakly-Supervised, Segmentation-Free, One-Step, Full Page Text Recognition by learning to unfoldCode1
Rescoring Sequence-to-Sequence Models for Text Line Recognition with CTC-PrefixesCode1
SmartPatch: Improving Handwritten Word Imitation with Patch DiscriminatorsCode1
TextAdaIN: Paying Attention to Shortcut Learning in Text RecognizersCode1
Exploring OCR Capabilities of GPT-4V(ision) : A Quantitative and In-depth EvaluationCode1
Best Practices for a Handwritten Text Recognition SystemCode1
BN-DRISHTI: Bangla Document Recognition through Instance-level Segmentation of Handwritten Text ImagesCode1
Hespi: A pipeline for automatically detecting information from hebarium specimen sheetsCode1
Sequence-to-Sequence Contrastive Learning for Text RecognitionCode1
Easter2.0: Improving convolutional models for handwritten text recognitionCode1
Enhancing Indic Handwritten Text Recognition Using Global Semantic InformationCode1
Classification of Handwritten Names of Cities and Handwritten Text Recognition using Various Deep Learning ModelsCode1
Manifold Mixup improves text recognition with CTC lossCode1
Muharaf: Manuscripts of Handwritten Arabic Dataset for Cursive Text RecognitionCode1
Continuous Offline Handwriting Recognition using Deep Learning ModelsCode1
KOHTD: Kazakh Offline Handwritten Text DatasetCode1
DAN: a Segmentation-free Document Attention Network for Handwritten Document RecognitionCode1
DANIEL: A fast Document Attention Network for Information Extraction and Labelling of handwritten documentsCode1
AttentionHTR: Handwritten Text Recognition Based on Attention Encoder-Decoder NetworksCode1
Boosting offline handwritten text recognition in historical documents with few labeled lines—0
A Scalable Handwritten Text Recognition System—0
Boosting Modern and Historical Handwritten Text Recognition with Deformable Convolutions—0
Books of Hours. the First Liturgical Data Set for Text Segmentation.—0
Are 2D-LSTM really dead for offline text recognition?—0
A limited-size ensemble of homogeneous CNN/LSTMs for high-performance word classification—0
Arabic Handwritten Text for Person Biometric Identification: A Deep Learning Approach—0
Align, Minimize and Diversify: A Source-Free Unsupervised Domain Adaptation Method for Handwritten Text Recognition—0
Advancing Offline Handwritten Text Recognition: A Systematic Review of Data Augmentation and Generation Techniques—0
End to End Recognition System for Recognizing Offline Unconstrained Vietnamese Handwriting—0
Benchmarking Large Language Models for Handwritten Text Recognition—0
Applications of Machine Learning in Document Digitisation—0
End-to-end information extraction in handwritten documents: Understanding Paris marriage records from 1880 to 1940—0
Show:102550
← PrevPage 1 of 3Next →
All datasetsIAMLAM(line-level)IAM(line-level)READ2016(line-level)BelfortREAD 2016BenthamDigital PeterHKRIAM-BIAM-DSaint Gall

Benchmark Results

#ModelMetricClaimedVerifiedStatus
1Transformer w/ CNNCER7.62—Unverified
2FPHR Paragraph Level (~145 dpi)CER6.7—Unverified
3Leaky LP CellCER6.6—Unverified
4FPHR+Aug Line Level (~145 dpi)CER6.5—Unverified
5Start, Follow, ReadCER6.4—Unverified
6Decouple Attention NetworkCER6.4—Unverified
7FPHR+Aug Paragraph Level (~145 dpi)CER6.3—Unverified
8Easter2.0CER6.21—Unverified
9HTR-VT(line-level)CER4.7—Unverified
10Transformer w/ CNN (+synth)CER4.67—Unverified
#ModelMetricClaimedVerifiedStatus
1GFCNTest CER5.2—Unverified
2TrOCRTest CER3.6—Unverified
3OrigamiNet-18Test CER3.1—Unverified
4OrigamiNet-12Test CER3.1—Unverified
5OrigamiNet-24Test CER3—Unverified
6HTR-VTTest CER2.8—Unverified
#ModelMetricClaimedVerifiedStatus
1GFCNTest CER8—Unverified
2OrigamiNet-12Test CER6—Unverified
3VANTest CER5—Unverified
4HTR-VTTest CER4.7—Unverified
5TrOCRTest CER3.4—Unverified
#ModelMetricClaimedVerifiedStatus
1CNN + BLSTMTest CER4.7—Unverified
2SpanTest CER4.6—Unverified
3DANTest CER4.1—Unverified
4VANTest CER4.1—Unverified
5HTR-VTTest CER3.9—Unverified
#ModelMetricClaimedVerifiedStatus
1PyLaia (human transcriptions + random split)CER (%)10.54—Unverified
2PyLaia (human transcriptions + agreement-based split)CER (%)5.57—Unverified
3PyLaia (rover consensus + agreement-based split)CER (%)4.95—Unverified
4PyLaia (all transcriptions + agreement-based split)CER (%)4.34—Unverified
#ModelMetricClaimedVerifiedStatus
1HTR-VT(line-level)CER (%)3.9—Unverified
2DANCER (%)3.22—Unverified
#ModelMetricClaimedVerifiedStatus
1StackMix+BlotsCER1.73—Unverified
#ModelMetricClaimedVerifiedStatus
1StackMix+BlotsCER2.5—Unverified
#ModelMetricClaimedVerifiedStatus
1StackMix+BlotsCER3.49—Unverified
#ModelMetricClaimedVerifiedStatus
1StackMix+BlotsCER3.77—Unverified
#ModelMetricClaimedVerifiedStatus
1StackMix+BlotsCER3.01—Unverified
#ModelMetricClaimedVerifiedStatus
1StackMix+BlotsCER3.65—Unverified
#ModelMetricClaimedVerifiedStatus
1DANCER (%)6.46—Unverified