Deep cartoon colorizer: An automatic approach for colorization of vintage cartoons

Chybicki, Mariusz; Kozakiewicz, Wiktor; Sielski, Dawid; Fabijańska, Anna

doi:10.1016/j.engappai.2019.02.006

Cited by 7 publications

(6 citation statements)

References 12 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…In particular, black and white image colourisation [22][23][24][25][26] is the most widely used. Chybicki et al [27] achieve a fully automatic method to colourise vintage cartoons using CNN model. The DeOldify project proposes a method for old photos colouring and restoration, also for old movie video colouring by using self-attention GAN [26] network.…”

Section: Image Colourisationmentioning

confidence: 99%

“…At this time, we begin the network training. When testing, we decolourise the testing data to get the FIGURE 5 Comparison results of ink painting with CycleGAN [8], UNIT [30], AGGAN [33], DCC [27] and our Flower-GAN. We can see that our method generates more accurate images and mitigates the problem of artefact grayscale images as input and recolourise them with the trained network.…”

Section: Comparison Experimentsmentioning

confidence: 99%

“…Comparison results of ink painting with CycleGAN [8], UNIT [30], AGGAN [33], DCC [27] and our Flower‐GAN. We can see that our method generates more accurate images and mitigates the problem of artefact…”

Section: Experiments and Analysismentioning

confidence: 99%

See 2 more Smart Citations

Multi‐style Chinese art painting generation of flowers

Tang

et al. 2020

IET Image Processing

View full text Add to dashboard Cite

With the proposal and development of Generative Adversarial Networks, the great achievements in the field of image generation are made. Meanwhile, many works related to the generation of painting art have also been derived. However, due to the difficulty of data collection and the fundamental challenge from freehand expressions, the generation of traditional Chinese painting is still far from being perfect. This paper specialises in Chinese art painting generation of flowers, which is important and classic, by deep learning method. First, an unpaired flowers paintings data set containing three classic Chinese painting style: line drawing, meticulous, and ink is constructed. Then, based on the collected dataset, a Flower‐Generative Adversarial Network framework to generate multi‐style Chinese art painting of flowers is proposed. The Flower‐Generative Adversarial Network, consisting of attention‐guided generators and discriminators, transfers the style among line drawing, meticulous, and ink by an adversarial training way. Moreover, in order to solve the problem of artefact and blur in image generation by existing methods, a new loss function called Multi‐Scale Structural Similarity to force the structure preservation is introduced. Extensive experiments show that the proposed Flower‐Generative Adversarial Network framework can produce better and multi‐style Chinese art painting of flowers than existing methods.

show abstract

Section: Image Colourisationmentioning

confidence: 99%

Section: Comparison Experimentsmentioning

confidence: 99%

See 1 more Smart Citation

Multi‐style Chinese art painting generation of flowers

Tang

et al. 2020

IET Image Processing

View full text Add to dashboard Cite

show abstract

“…Os primeiros trabalhos sobre colorizadores automáticos implementados com Deep Learning datam de 2015. A abordagem típica consiste em fornecer uma imagem em escala de cinza como entrada de uma CNN e obter, em sua saída, matrizes referentes aos seus canais de cores [3], [6], [1], [7]. Abordagens alternativas envolvem a utilização de CNNs para produzir distribuições probabilísticas da colorização para cada pixel [4], [2], [8], [9], [10], ou o uso de Redes Generativas Adversariais (GANs [11]) para produzir um colorizador a partir da competição entre duas redes neurais [12], [13].…”

Section: A Colorização Automática Com Aprendizagem Profundaunclassified

“…Os canais de cores são usualmente representados nos espaços de cores CIE Lab [4], [2], [3], [1], [7], [9], [10], YUV [6], [8], [12], ou HCL [3]. A vantagem dessas alternativas é que nesses espaços a imagem em escala de cinza corresponde a um dos canais da imagem colorida, diferentemente do espaço RGB, não sendo necessário que a rede produza informações redundantes.…”

Section: A Colorização Automática Com Aprendizagem Profundaunclassified

Colorização Automática de Imagens em Escala de Cinza Utilizando Aprendizagem Profunda

Araújo¹,

Veloso²

2020

Anais De XXXVIII Simpósio Brasileiro De Telecomunicações E Processamento De Sinais

View full text Add to dashboard Cite

Resumo-Métodos de Aprendizagem Profunda (Deep Learning) têm superado técnicas tradicionais em diversas aplicações de domínios da Visão Computacional e do Processamento de Sinais. Nesse contexto, esse trabalho realiza um estudo sobre a aplicabilidade de tais métodos para colorir imagens automaticamente. Três arquiteturas de Redes Neurais Convolucionais foram comparadas segundo três índices quantitativos, a raíz do erro médio quadrático (RMSE), a relação sinal ruído de pico (PSNR) e o índice de similaridade estrutural (SSIM), além de um teste de usabilidade. Foi verificado que os modelos avaliados são capazes de realizar a colorização automática, sendo capazes de produzir resultados fotorrealistas.

show abstract

Image colorization using deep convolutional auto-encoder with multi-skip connections

Jin

Jiang

et al. 2022

Soft Comput

View full text Add to dashboard Cite

Deep cartoon colorizer: An automatic approach for colorization of vintage cartoons

Cited by 7 publications

References 12 publications

Multi‐style Chinese art painting generation of flowers

Multi‐style Chinese art painting generation of flowers

Colorização Automática de Imagens em Escala de Cinza Utilizando Aprendizagem Profunda

Image colorization using deep convolutional auto-encoder with multi-skip connections

Contact Info

Product

Resources

About