1,291 Hits in 4.8 sec

Adversarial Generation of Handwritten Text Images Conditioned on Sequences [article]

Eloi Alonso, Bastien Moysset, Ronaldo Messina
2019 arXiv   pre-print
In order to partially satisfy this requirement, we propose a system based on Generative Adversarial Networks (GAN) to produce synthetic images of handwritten words.  ...  We obtain realistic images on both French and Arabic datasets, and we show that integrating these synthetic images into the existing training data of a text recognition system can slightly enhance its  ...  CONCLUSION We presented an adversarial model to produce synthetic images of handwritten word images, conditioned on the sequence of characters to render.  ... 
arXiv:1903.00277v1 fatcat:y6xjm5uxhreqjow4as7evvle4y

ThaiWritableGAN: Handwriting Generation under Given Information

Lawankorn Mookdarsanit, Pakpoom Mookdarsanit
2021 International Journal of Computing and Digital Systems  
D is assigned to discriminate an unknown handwritten image that it is real or generated.  ...  R is a convolutional neural network (pre-trained by real Thai handwritten images) that is additionally added to recognize the synthesized images (generated by G).  ...  A type of conditional GAN for imageto-image translation (the well-known name as "pix2pix") [47] was used to generate many images in 2017.  ... 
doi:10.12785/ijcds/100165 fatcat:v46prnxovbdspplgugy5o2plge

GANwriting: Content-Conditioned Generation of Styled Handwritten Word Images [article]

Lei Kang, Pau Riba, Yaxing Wang, Marçal Rusiñol, Alicia Fornés, Mauricio Villegas
2020 arXiv   pre-print
Although current image generation methods have reached impressive quality levels, they are still unable to produce plausible yet diverse images of handwritten words.  ...  We propose a novel method that is able to produce credible handwritten word images by conditioning the generative process with both calligraphic style features and textual content.  ...  Yet, for the specific case of generating handwritten text, one could also envisage the option of directly producing the final images instead of generating the stroke sequences needed to pencil a particular  ... 
arXiv:2003.02567v2 fatcat:ooztnppezraltfgdczankeryru

Diffusion models for Handwriting Generation [article]

Troy Luhman, Eric Luhman
2020 arXiv   pre-print
Experiments reveal that our model is able to generate realistic , high quality images of handwritten text in a similar style to a given writer.  ...  Our method of handwriting generation does not require using any text-recognition based, writer-style based, or adversarial loss functions, nor does it require training of auxiliary networks.  ...  Each sample of handwriting data is associated with a text sequence label describing the content of the handwritten text.  ... 
arXiv:2011.06704v1 fatcat:2mgqwpkqbrab3g44uawh3ojqaa

Handwriting Transformers [article]

Ankan Kumar Bhunia, Salman Khan, Hisham Cholakkal, Rao Muhammad Anwer, Fahad Shahbaz Khan, Mubarak Shah
2021 arXiv   pre-print
To the best of our knowledge, we are the first to introduce a transformer-based generative network for styled handwritten text generation.  ...  Our proposed HWT generates realistic styled handwritten text images and significantly outperforms the state-of-the-art demonstrated through extensive qualitative, quantitative and human-based evaluations  ...  [3] propose an approach, where handwritten text generation is conditioned by character sequences. However, their approach suffers from style collapse hindering the diversity of synthesized images.  ... 
arXiv:2104.03964v1 fatcat:6dveynb6hrhplevpia4mpgeb2m

Unsupervised Training Data Generation of Handwritten Formulas using Generative Adversarial Networks with Self-Attention [article]

Matthias Springstein and Eric Müller-Budack and Ralph Ewerth
2021 arXiv   pre-print
For this purpose, we propose a novel attention-based generative adversarial network to translate rendered equations to handwritten formulas.  ...  The recognition of handwritten mathematical expressions in images and video frames is a difficult and unsolved problem yet.  ...  Another way to create handwritten text is to generate the strokes of the individual symbols instead of calculating the image pixel by pixel.  ... 
arXiv:2106.09432v1 fatcat:2j5wi7e2x5hyjmh3ygf2673hfi

Illegible Text to Readable Text: An Image-to-Image Transformation using Conditional Sliced Wasserstein Adversarial Networks [article]

Mostafa Karimi, Gopalkrishna Veni, Yen-Yun Yu
2019 arXiv   pre-print
We tackle this problem by developing a handwritten-to-machine-print conditional Generative Adversarial network (HW2MP-GAN) model that formulates handwritten recognition as a text-Image-to-text-Image translation  ...  for image-to-image translation on IAM database.  ...  Our results on IAM database reveal the superiority of the proposed model on state-of-the-art conditional GAN models for handwritten image to machine-print image translation.  ... 
arXiv:1910.05425v1 fatcat:k7vbel5dungdfn7qws7qbeuswy

Enhance to Read Better: A Multi-Task Adversarial Network for Handwritten Document Image Enhancement [article]

Sana Khamekhem Jemni and Mohamed Ali Souibgui and Yousri Kessentini and Alicia Fornés
2021 arXiv   pre-print
In this paper, we propose an end to end architecture based on Generative Adversarial Networks (GANs) to recover the degraded documents into a clean and readable form.  ...  the generated document image to be more readable.  ...  The first one was conditioning on the clean image to generate a degraded one, while the other network reconstructed the clean version conditioning on the degraded image.  ... 
arXiv:2105.12710v2 fatcat:oqri452ocjeqraj4g6t2yxjrpu

SLOGAN: Handwriting Style Synthesis for Arbitrary-Length and Out-of-Vocabulary Text [article]

Canjie Luo, Yuanzhi Zhu, Lianwen Jin, Zhe Li, Dezhi Peng
2022 arXiv   pre-print
To this end, we propose a novel method that can synthesize parameterized and controllable handwriting Styles for arbitrary-Length and Out-of-vocabulary text based on a Generative Adversarial Network (GAN  ...  Moreover, we embed the text content by providing an easily obtainable printed style image, so that the diversity of the content can be flexibly achieved by changing the input printed image.  ...  a text string image [13] , [15] , [21] , because one text string image contains multiple objects in a sequence [22] , [23] .  ... 
arXiv:2202.11456v1 fatcat:c6vpdg2v4rge5iymjfnk3rfyb4

Deep Reader: Information extraction from Document images via relation extraction and Natural Language [article]

Vishwanath D, Rohit Rahul, Gunjan Sehgal, Swati, Arindam Chowdhury, Monika Sharma, Lovekesh Vig, Gautam Shroff, Ashwin Srinivasan
2018 arXiv   pre-print
DeepReader has a suite of state-of-the-art vision algorithms which are applied to recognize handwritten and printed text, eliminate noisy effects, identify the type of documents and detect visual entities  ...  In this paper, we also demonstrate many different capabilities of Deep Reader and report results on a real-world use case.  ...  These include utilization of generative adversarial networks [12] for image denoising, Siamese networks [13] for document template identification and sequence to sequence models for handwritten text  ... 
arXiv:1812.04377v2 fatcat:mtot4zi3lbe5xkmppiw3eeafru

Realistic Handwriting Generation using Generative Adversarial Networks (Rnn)

2020 International journal of recent technology and engineering  
It will then be further drawn out to handwriting generation. The created network will be conditioning its predictions based on a sequence of text.  ...  Here, we will be trying to predict one point of data at a time. Our approach is shownfor text, where the type of data is discrete.  ...  It will then be further drawn out to handwriting generation. The created network will be conditioning its predictions based on a sequence of text.  ... 
doi:10.35940/ijrte.a2761.059120 fatcat:hutfek64effe5geucqoowz7kgm

SmartPatch: Improving Handwritten Word Imitation with Patch Discriminators [article]

Alexander Mattick, Martin Mayr, Mathias Seuret, Andreas Maier, Vincent Christlein
2021 arXiv   pre-print
As of recent generative adversarial networks have allowed for big leaps in the realism of generated images in diverse domains, not the least of which being handwritten text generation.  ...  The generation of realistic-looking hand-written text is important because it can be used for data augmentation in handwritten text recognition (HTR) systems or human-computer interaction.  ...  Handwritten Text Recognition We evaluate the legibility of the words by training a state-of-the-art handwritten text recognition model [14] on the IAM dataset [15] and by performing evaluations on  ... 
arXiv:2105.10528v1 fatcat:kmma2jkbqfe5xotedeu5pupy2a

Offline Hand Written Urdu Word Spotting using Random Data Generation

Faiq Faizan Farooqui, Muhammad Hassan, Muhammad Shahzad Younis, Muhammad Kashif Siddhu
2020 IEEE Access  
For image generation, we used geometric transformations and variants of Generative Adversarial Network (GAN).  ...  Urdu word spotting is among the most challenging tasks in image processing and word spotting of hand written Urdu text is even more so.  ...  GANs were used to generate handwritten text by Wang et al. [38] .  ... 
doi:10.1109/access.2020.3010166 fatcat:ybwmfq7dfnecvkiz7czszbnd5y

TextAdaIN: Paying Attention to Shortcut Learning in Text Recognizers [article]

Oren Nuriel, Sharon Fogel, Ron Litman
2021 arXiv   pre-print
TextAdaIN achieves state-of-the-art results on standard handwritten text recognition benchmarks. Additionally, it generalizes to multiple architectures and to the domain of scene text recognition.  ...  However in some cases, their decisions are based on unintended information leading to high performance on standard benchmarks but also to a lack of generalization to challenging testing conditions and  ...  Our method achieves state-of-the-art results on handwritten text recognition benchmarks and improves robustness towards challenging testing conditions.  ... 
arXiv:2105.03906v2 fatcat:v4pel3x4d5cttiivvrjdccgsgi

Text and Style Conditioned GAN for Generation of Offline Handwriting Lines [article]

Brian Davis, Chris Tensmeyer, Brian Price, Curtis Wigington, Bryan Morse, Rajiv Jain
2020 arXiv   pre-print
This paper presents a GAN for generating images of handwritten lines conditioned on arbitrary text and latent style vectors.  ...  Unlike prior work, which produce stroke points or single-word images, this model generates entire lines of offline handwriting.  ...  Introduction In this work, we generate images of lines of handwriting, conditioned on the desired text and a latent style vector.  ... 
arXiv:2009.00678v1 fatcat:4aj7ut4jpnau5oeiaqach6ydoa
« Previous Showing results 1 — 15 out of 1,291 results