SOTAVerified

Optical Character Recognition (OCR)

Optical Character Recognition or Optical Character Reader (OCR) is the electronic or mechanical conversion of images of typed, handwritten or printed text into machine-encoded text, whether from a scanned document, a photo of a document, a scene-photo (for example the text on signs and billboards in a landscape photo, license plates in cars...) or from subtitle text superimposed on an image (for example: from a television broadcast)

Papers

Recently AddedMost HypedMost ActiveNeeds VerificationMost Verified

Showing 126–150 of 1209 papers

TitleStatusHype
Awaker2.5-VL: Stably Scaling MLLMs with Parameter-Efficient Mixture of ExpertsCode1
FigStep: Jailbreaking Large Vision-Language Models via Typographic Visual PromptsCode1
End-to-End Information Extraction by Character-Level Embedding and Multi-Stage Attentional U-NetCode1
Easter2.0: Improving convolutional models for handwritten text recognitionCode1
Enhancing License Plate Super-Resolution: A Layout-Aware and Character-Driven ApproachCode1
Attack of the Tails: Yes, You Really Can Backdoor Federated LearningCode1
AT-ST: Self-Training Adaptation Strategy for OCR in Domains with Limited TranscriptionsCode1
EAST: An Efficient and Accurate Scene Text DetectorCode1
Exploring Better Text Image Translation with Multimodal CodebookCode1
DocReal: Robust Document Dewarping of Real-Life Images via Attention-Enhanced Control Point PredictionCode1
DocFormerv2: Local Features for Document UnderstandingCode1
Fully Unsupervised Diversity Denoising with Convolutional Variational AutoencodersCode1
DocLayLLM: An Efficient and Effective Multi-modal Extension of Large Language Models for Text-rich Document UnderstandingCode1
DocScanner: Robust Document Image Rectification with Progressive LearningCode1
Detection of Furigana Text in ImagesCode1
DocLayLLM: An Efficient Multi-modal Extension of Large Language Models for Text-rich Document UnderstandingCode1
DocParser: End-to-end OCR-free Information Extraction from Visually Rich DocumentsCode1
A Multiplexed Network for End-to-End, Multilingual OCRCode1
Digitizing Historical Balance Sheet Data: A Practitioner's GuideCode1
DE-GAN: A Conditional Generative Adversarial Network for Document EnhancementCode1
DSG: An End-to-End Document Structure GeneratorCode1
DiT: Self-supervised Pre-training for Document Image TransformerCode1
A Two-Step Approach for Automatic OCR Post-CorrectionCode1
Efficient OCR for Building a Diverse Digital HistoryCode1
DocTr: Document Image Transformer for Geometric Unwarping and Illumination CorrectionCode1
Show:102550
← PrevPage 6 of 49Next →
All datasetsBenchmarking Chinese Text Recognition: Datasets, Baselines, and an Empirical StudyVideoDB's OCR Benchmark Public CollectionFSNS - TestI2L-140KSUTim2latex-100k

Benchmark Results

#ModelMetricClaimedVerifiedStatus
1DTrOCRAccuracy (%)89.6—Unverified
2DTrOCR 105MAccuracy (%)89.6—Unverified
3MaskOCR-LAccuracy (%)82.6—Unverified
4TransOCRAccuracy (%)72.8—Unverified
5SRNAccuracy (%)65—Unverified
6MORANAccuracy (%)64.3—Unverified
7SEEDAccuracy (%)61.2—Unverified
#ModelMetricClaimedVerifiedStatus
1GPT-4oAverage Accuracy76.22—Unverified
2Gemini-1.5 ProAverage Accuracy76.13—Unverified
3Claude-3 SonnetAverage Accuracy67.71—Unverified
4RapidOCRAverage Accuracy56.98—Unverified
5EasyOCRAverage Accuracy49.3—Unverified
#ModelMetricClaimedVerifiedStatus
1STREETSequence error27.54—Unverified
2SEESequence error22—Unverified
3AttentionOCR_Inception-resnet-v2_LocationSequence error15.8—Unverified
#ModelMetricClaimedVerifiedStatus
1I2L-NOPOOLBLEU89.09—Unverified
2I2L-STRIPSBLEU89—Unverified
#ModelMetricClaimedVerifiedStatus
1TesseractCharacter Error Rate (CER)0.08—Unverified
2EasyOCRCharacter Error Rate (CER)0.07—Unverified
#ModelMetricClaimedVerifiedStatus
1I2L-STRIPSBLEU88.86—Unverified