Employing Neural Style Transfer for Generating Deep Dream Images

Keywords: Deep dream, Gradient ascent, Gram matrix, Neural style transfer

Abstract

In recent years, deep dream and neural style transfer emerged as hot topics in deep learning. Hence, mixing those two techniques support the art and enhance the images that simulate hallucinations among psychiatric patients and drug addicts. In this study, our model combines deep dream and neural style transfer (NST) to produce a new image that combines the two technologies. VGG-19 and Inception v3 pre-trained networks are used for NST and deep dream, respectively. Gram matrix is a vital process for style transfer. The loss is minimized in style transfer while maximized in a deep dream using gradient descent for the first case and gradient ascent for the second. We found that different image produces different loss values depending on the degree of clarity of that images. Distorted images have higher loss values in NST and lower loss values with deep dreams. The opposite happened for the clear images that did not contain mixed lines, circles, colors, or other shapes.

Downloads

Download data is not yet available.

Author Biographies

Lafta R. Al-Khazraji, (1) Department of Computer Science, University of Technology-Iraq, Baghdad, Iraq. (2) General Directorate of Education of Salahuddin Governorate, Iraq.

Lafta R. Al-Khazraji is an Assistant Lecturer at the General Directorate of Education of Salahuddin Governorate, Iraq. He received the B.Sc. degree in Computer Science from Al Rafidain University College, the M.Sc. degree in Computer Science from Iraqi Commission for Computers & Informatics. Recently, he is a Ph.D. candidate at the department of Computer Science, University of Technology. His research interests include artificial intelligence, image processing, machin learning, deep learning, pattern recognition and computer networks.

Ayad R. Abbas, Department of Computer Science, University of Technology-Iraq, Baghdad, Iraq.

Ayad R. Abbas is a Professor at the Computer Science Department, University of Technology, Iraq. He has a Ph.D. degree in Applied Computer Technology from Wuhan University, School of Computer Science, China. His research interests include artificial intelligent, machine learning, natural language processing, deep learning, data mining, web mining, information retrieval, soft computing, E-learning, E-commerce and recommended systems.

Abeer S. Jamil, Department of Computer Technology Engineering, Al-Mansour University College, Baghdad, Iraq

Abeer S. Jamil received the MSc. and Ph.D. degree in Computer Science from the University of Technology, Iraq. Her research interests are digital image processing, video processing, security software engineering, networking, and artificial intelligence applications.

References

Abedi, W.M., Nadher, I., Sadiq, A.T. and Al., E., 2020. Modified deep learning method for body postures recognition. International Journal of Advanced Science and Technology, 29, pp.3830-3841.

Ali, L., Alnajjar, F., Jassmi, H.A, Gochoo, M., Khan, W. and Serhani, M.A., 2021. Performance evaluation of deep CNN-based crack detection and localization techniques for concrete structures. Sensors, 21, p.1688.

Alzubaidi, L., Zhang, J., Humaidi, A.J., Al-Dujaili, A., Duan, Y., Al-Shamma, O., Santamaría, J., Fadhel, M.A., Al-Amidie, M. and Farhan, L., 2021. Review of deep learning: Concepts, CNN architectures, challenges, applications, future directions. Journal of Big Data, 8, pp.1-74.

Cao, J., Yan, M., Jia, Y., Tian, X. and Zhang, Z., 2021. Application of a modified Inception-v3 model in the dynasty-based classification of ancient murals. EURASIP Journal on Advances in Signal Processing, 2021, p.49.

Chen, X., Zhang, Y., Wang, Y., Shu, H., Xu, C. and Xu, C., 2020. Optical flow distillation: Towards efficient and stable video style transfer. In: Lecture Notes in Computer Science (LNCS). Springer Science, Germany.

Choi, H.C., 2022. Toward exploiting second-order feature statistics for arbitrary image style transfer. Sensors(Basel), 2022, p.2611.

El-Rahiem, B.A., Amin, M., Sedik, A., Samie, F.E. and Iliyasu, A.M., 2022. An efficient multi-biometric cancellable biometric scheme based on deep fusion and deep dream. Journal of Ambient Intelligence and Humanized Computing, 13, pp.2177-2189.

Gatys, L.A., Ecker, A.S. and Bethge, M., 2015. A Neural Algorithm of Artistic Style. arXiv Prepr. arXiv1508.06576.

Khan, A., Sohail, A., Zahoora, U. and Qureshi, A.S., 2020. A survey of the recent architectures of deep convolutional neural networks. Artificial Intelligence Review, 53, pp.5455-5516.

Kiran, T.T., 2021. Deep inceptionism learning performance analysis using TensorFlow with GPU-deep dream algorithm. Journal of Emerging Technologies and Innovative Research, 8, pp.322-328.

Kotovenko, D., Sanakoyeu, A., Ma, P., Lang, S. and Ommer, B., 2019. A content transformation block for image style transfer. In: Proceedings of the IEEE Computer Society Conference on Computer Vision and Pattern Recognition. IEEE, United States, pp. 10024-10033.

Li, H., 2018. A Literature Review of Neural Style Transfer. Princeton University Technical Report, Princeton NJ, p.085442019.

Mordvintsev, A., Olah, C. and Tyka, M., 2015. Inceptionism: Going Deeper into Neural Networks. Available from : https://googleresearch.blogspot.co.uk/2015/06/ inceptionism-going-deeper-into-neural.html [Last accessed on 2022 Aug 03].

Rashid, M., Khan, M.A., Alhaisoni, M., Wang, S.H., Naqvi, S.R., Rehman, A. and Saba, T., 2020. A sustainable deep learning framework for object recognition using multi-layers deep features fusion and selection. Sustain, 12, p.1-21.

Singh, A., Jaiswal, V., Joshi, G., Sanjeeve, A., Gite, S. and Kotecha, K., 2021. Neural style transfer: A critical review. IEEE Access, 9, pp.131583-131613.

Sudha, V. and Ganeshbabu, T.R., 2021. A convolutional neural network classifier VGG-19 architecture for lesion detection and grading in diabetic retinopathy based on deep learning. Computers, Materials and Continua, 66, pp.827-842.

Szegedy, C., Vanhoucke, V., Ioffe, S., Shlens, J. and Wojna, Z., 2016. Rethinking the inception architecture for computer vision. In: Proceedings of the IEEE Computer Society Conference on Computer Vision and Pattern Recognition. IEEE, United States, pp.2818-2826.

Tuyen, N.Q., Nguyen, S.T., Choi, T.J. and Dinh, V.Q., 2021. Deep correlation multimodal neural style transfer. IEEE Access, 9, p.141329-141338.

Wani, M.A., Bhat, F.A., Afzal, S. and Khan, A.I., 2020. Advances in Deep Learning. Springer Nature, Singapore.

Xiao, J., Wang, J., Cao, S. and Li, B., 2020. Application of a novel and improved VGG-19 network in the detection of workers wearing masks. Journal of Physics: Conference Series, 1518, 012041.

Yin, H., Molchanov, P., Alvarez, J.M., Li, Z., Mallya, A., Hoiem, D., Jha, N.K. and Kautz, J., 2020. Dreaming to distill: Data-free knowledge transfer via deepinversion. In: 2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR). IEEE, United States, pp.8712-8721.

Published
2022-12-01
How to Cite
Al-Khazraji, L. R., Abbas, A. R. and Jamil, A. S. (2022) “Employing Neural Style Transfer for Generating Deep Dream Images”, ARO-THE SCIENTIFIC JOURNAL OF KOYA UNIVERSITY, 10(2), pp. 134-141. doi: 10.14500/aro.11051.