Manipulasi Gambar dengan Transfer Gaya Menggunakan Convolutional Neural Network
-
Rakhmi Khalida
*
Universitas Bhayangkara Jakarta Raya, Bekasi,
Indonesia
- Khairunnisa Fadhilla Ramdhania Universitas Bhayangkara Jakarta Raya, Bekasi, Indonesia (*) Corresponding Author
Abstract
Recently computers have been able to produce photographs that allow users to compose selfies with van Gogh paintings. Inspired by the power of convolutional neural networks (CNN), he first learned how to use CNN to reproduce famous painting styles combined with self-portrait images. The method used is called a neural network transfer. However, early versions of neural networks had optimization problems, requiring hundreds or thousands of iterations to transfer forces combined with a single image. To overcome this in-efficiency, researchers developed the CNS-style PerStyle-Per-Model (PSPM) transfer method. The development of force transfer using a deep neural network is also called NST by training the VGG-16 model to change any image in one feed, foward propagation. A trained model can adjust to any drawing mode with just one iteration instead of thousands of iterations over the network and to get the most objective possible style of transfer
Downloads
References
B. Gooch and A. Gooch, Non photorealistic rendering. Natick. MA, USA: A. K. Peters, Ltd, 2001.
T. Strothotte and S. Schlechtweg, Non-photorealistic computer graphics: modeling, rendering, and animation. Morgan Kaufmann, 2002.
P. Rosin and J. Collomosse, “Image and video-based artistic stylisation,” Springer Sci. Bus. Media, vol. 42, 2012.
and C. V. S. L. A. Gatys, A. S. Ecker, M. Bethge, “A Neural Algorithm of Artistic Style,” pp. 3–7, 2015.
and V. L. D. Ulyanov, V. Lebedev, A. Vedaldi, “Texture networks: Feed-forward synthesis of textures and stylized images,” Int. Conf. Mach. Learn., pp. 1349–1357, 2016.
and L. F. J. Johnson, A. Alahi, “Perceptual Losses for Real-Time Style Transfer and Super-Resolution,” arXiv Prepr., 2016.
C. Li and M. Wand, “Precomputed real-time texture synthesis with markovian generative adversarial networks,” Eur. Conf. Comput. Vis., pp. 702–716, 2016.
and M. T. A. Mordvintsev, C. Olah, “Inceptionism: Going deeper into neural networks,” 2015. [Online]. Available: https://research.googleblog.com/2015/06/ inceptionism-going-deeper-into-neural.html.
and L. M. H. Huang, H. Wang, W. Luo, “Real-Time Neural Style Transfer for Videos,” pp. 783–791.
and B. S. G. Amy A. Gooch, Jeremy Long, Li Ji, Anthony Estey, “Viewing Progress in Non-Photorealistic Rendering Through Heinlein’s Lens,” in In Proc. NPAR. ACM, New York, 2010, pp. 165–171.
and T. S. Nick Halper, Mara Mellin, Christoph S. Herrmann, Volker Linneweber, “Psychology and Non-Photorealistic Rendering,” Begin. a Beautiful Relationship. Proc. Mensch Comput., pp. 277–286, 2003.
David H. Salesin, “Non-Photorealistic Animation & Rendering,” in NPAR, 2002.
and E. R. Hasan Sheikh Faridul, Tania Pouli, Christel Chamaret, Jürgen Stauder, Alain Trémeau, “A Survey of Color Mapping and its Applications,” 2014.
I. P. Labs, “Prisma: Turn memories into art using artificial intelligence,” 2016. .
Ostagram, “Ostagram,” 2016. .
T. Henighan, “Spatial Control in Neural Style Transfer,” 2017.
G. Pan, D. Sun, R. Zhan, and J. Zhang, “Mural Sketch Generation via Style-aware Convolutional Neural Network,” in CGI 2018: Proceedings of Computer Graphics International 2018, 2018, pp. 239–245.
D. J. Semmo, Amir, Isenberg Tobias, “Neural Style Transfer : A Paradigm Shift for Image-based Artistic Rendering,” in Proceedings of the Symposium on Non-Photorealistic Animation and Rendering, 2017.
Bila bermanfaat silahkan share artikel ini
Berikan Komentar Anda terhadap artikel Manipulasi Gambar dengan Transfer Gaya Menggunakan Convolutional Neural Network
Pages: 244-252
Copyright (c) 2021 Rakhmi Khalida, Khairunnisa Fadhilla Ramdhania
This work is licensed under a Creative Commons Attribution 4.0 International License.
Authors who publish with this journal agree to the following terms:
- Authors retain copyright and grant the journal right of first publication with the work simultaneously licensed under Creative Commons Attribution 4.0 International License that allows others to share the work with an acknowledgment of the work's authorship and initial publication in this journal.
- Authors are able to enter into separate, additional contractual arrangements for the non-exclusive distribution of the journal's published version of the work (e.g., post it to an institutional repository or publish it in a book), with an acknowledgment of its initial publication in this journal.
- Authors are permitted and encouraged to post their work online (e.g., in institutional repositories or on their website) prior to and during the submission process, as it can lead to productive exchanges, as well as earlier and greater citation of published work (Refer to The Effect of Open Access).
