Handwritten Text Recognition
37 papers with code • 9 benchmarks • 10 datasets
Benchmarks
These leaderboards are used to track progress in Handwritten Text Recognition
| Trend | Dataset | Best Model | Paper | Code | Compare |
|---|---|---|---|---|---|
|
|||||
|
|||||
|
|||||
|
|||||
|
|||||
|
|||||
|
|||||
|
|||||
|
Datasets
IAM
READ 2016
HKR
Bentham
Digital Peter
Saint Gall
100 People - Handwriting OCR Data of Japanese and Korean
SIMARA
Most implemented papers
Full Page Handwriting Recognition via Image to Sequence Extraction
kingyiusuen/image-to-latex
•
•
We present a Neural Network based Handwritten Text Recognition (HTR) model architecture that can be trained to recognize full pages of handwritten or printed text without image segmentation.
Decoupled Attention Network for Text Recognition
Canjie-Luo/Scene-Text-Image-Transformer
•
•
To remedy this issue, we propose a decoupled attention network (DAN), which decouples the alignment operation from using historical decoding results.
ScrabbleGAN: Semi-Supervised Varying Length Handwritten Text Generation
amzn/convolutional-handwriting-gan
•
•
CVPR 2020
This is especially true for handwritten text recognition (HTR), where each author has a unique style, unlike printed text, where the variation is smaller by design.
Sequence-to-Sequence Contrastive Learning for Text Recognition
amazon-science/semimtr-text-recognition
•
•
CVPR 2021
We propose a framework for sequence-to-sequence contrastive learning (SeqCLR) of visual representations, which we apply to text recognition.
Digital Peter: Dataset, Competition and Handwriting Recognition Methods
This paper presents a new dataset of Peter the Great's manuscripts and describes a segmentation procedure that converts initial images of documents into the lines.
TrOCR: Transformer-based Optical Character Recognition with Pre-trained Models
microsoft/unilm
•
•
Text recognition is a long-standing research problem for document digitalization.
Character-Based Handwritten Text Transcription with Attention Networks
When the sequence alignment is one-to-one, softmax attention is able to learn a more precise alignment at each step of the decoding, whereas the alignment generated by sigmoid attention is much less precise.
Start, Follow, Read: End-to-End Full-Page Handwriting Recognition
cwig/start_follow_read
•
•
ECCV 2018
Despite decades of research, offline handwriting recognition (HWR) of degraded historical documents remains a challenging problem, which if solved could greatly improve the searchability of online cultural heritage archives.
No Padding Please: Efficient Neural Handwriting Recognition
Neural handwriting recognition (NHR) is the recognition of handwritten text with deep learning models, such as multi-dimensional long short-term memory (MDLSTM) recurrent neural networks.
Manifold Mixup improves text recognition with CTC loss
simplify23/Ultra_light_OCR_No.11
•
•
Modern handwritten text recognition techniques employ deep recurrent neural networks.
