Abstract
Image style transfer based on deep learning is an image processing method, which uses deep convolutional neural networks (CNN) to yield artistic images with some specific styles by segregating and restructuring the styles and contents of images [1]. Considering that several drawbacks still exists in the neural style transfer, such as the intrinsic limitation of single style transfer and the time-consuming process of optimizing model. At this time, we could remedy aforesaid weaknesses by the means of making use of fast multi-style transfer and its image quality assessment (IQA). On one hand, it aims to solve the monotony of color and the loss of stereoscopy results from single style image. On the other hand, it evaluates the synthesized image quality in real time, and makes the feedback of IQA to model itself for modifying relevant parameters of the algorithm, providing reference for selecting appropriate time to stop iterating to save immediate model. In this paper, we present the improved approaches for image style transfer to obtain preferable visual effect and save significant time.
This is a preview of subscription content, log in via an institution.
Buying options
Tax calculation will be finalised at checkout
Purchases are for personal use only
Learn about institutional subscriptionsReferences
Li Y, Zhang T, Han X, Qi Y (2018) Image style transfer in deep learning networks. In: 2018 5th international conference on systems and informatics (ICSAI)
Gatys LA, Ecker AS, Bethge M (2015) A neural algorithm of artistic style. arXiv preprint arXiv:1508.06576
Johnson J, Alahi A, Fei-Fei L (2016) Perceptual losses for real-time style transfer and super-resolution
Zhang W, Borji A, Wang Z et al (2016) The application of visual saliency models in objective image quality assessment: a statistical evaluation. IEEE Trans Neural Netw Learn Syst 27(6):1266–1278
Simonyan K, Zisserman A (2014) Very deep convolutional networks for large-scale image recognition. arXiv preprint arXiv:1409.1556
Bryan B, Raymond P, Andreas S (2018) Faster Art-CNN: an extremely fast style transfer network. In: IEEE western New York image and signal processing workshop
Wang Z, Bovik AC, Lu L (2011) Why is image quality assessment so difficult? In: IEEE international conference on Acoustics IEEE
Bovik AC (2013) Automatic prediction of perceptual image and video quality. Proc IEEE 99:1–17
Srivastava N, Hinton G, Krizhevsky A et al (2014) Dropout: a simple way to prevent neural networks from overfitting. J Mach Learn Res 15(1):1929–1958
Montavon G, Samek W, Müller K-R (2017) Methods for interpreting and understanding deep neural networks. Digit Sig Process 73:1–15
Bosse S, Siekmann M, Wiegand T, Samek W (2017) A perceptually relevant shearlet-based adaptation of the PSNR. In: IEEE international conference image processing (ICIP), pp 315–319
Soundararajan R, Bovik AC (2012) RRED indices: reduced reference entropic differencing for image quality assessment. IEEE Trans Image Process 21(2):517–526
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2020 Springer Nature Singapore Pte Ltd.
About this paper
Cite this paper
Zhao, X., Gao, H. (2020). Fast Image Multi-style Transfer and Its Quality Assessment. In: Jia, Y., Du, J., Zhang, W. (eds) Proceedings of 2019 Chinese Intelligent Systems Conference. CISC 2019. Lecture Notes in Electrical Engineering, vol 594. Springer, Singapore. https://doi.org/10.1007/978-981-32-9698-5_44
Download citation
DOI: https://doi.org/10.1007/978-981-32-9698-5_44
Published:
Publisher Name: Springer, Singapore
Print ISBN: 978-981-32-9697-8
Online ISBN: 978-981-32-9698-5
eBook Packages: Intelligent Technologies and RoboticsIntelligent Technologies and Robotics (R0)