Deep Network Interpolation for Continuous Imagery Effect Transition
By Xintao Wang, K. Yu, Chao Dong, Xiaoou Tang, Chen Change Loy. 2018
113 citations recorded by the source corpus.
Deep convolutional neural network has demonstrated its capability of learning a deterministic mapping for the desired imagery effect. However, the large variety of user flavors motivates the possibility of continuous transition among different output effects. Unlike existing methods that require a specific design to achieve one particular transition (e.g., style transfer), we propose a simple yet universal approach to attain a smooth control of diverse imagery effects in many low-level vision tasks, including image restoration, image-to-image translation, and style transfer. Specifically, our method, namely Deep Network Interpolation (DNI), applies linear interpolation in the parameter space of two or more correlated networks. A smooth control of imagery effects can be achieved by tweaking the interpolation coefficients. In addition to DNI and its broad applications, we also investigate the mechanism of network interpolation from the perspective of learned filters.
Subfields: Computer Vision and Pattern Recognition, Generative Models
How researchers see this paper
The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.
In Computer Vision and Pattern Recognition
- Chen Change Loy
- Luc Van Gool
- Ming-Hsuan Yang
- Xiaogang Wang
- Dacheng Tao
- Daniel Cohen-Or
- Andrea Vedaldi
- Trevor Darrell
In Generative Models
- Daniel Cohen-Or
- Or Patashnik
- Amit H. Bermano
- Rinon Gal
- Yujun Shen
- Eli Shechtman
- Chen Change Loy
- Bolei Zhou
Citations in the corpus
- StyleGAN-NADA (2021, methodology): “For more applications of network interpolations, see [54] or the concurrent work by Wu et al .”
- InterFaceGAN: Interpreting the Disentangled Face Representation Learned by GANs (2020, methodology): “(iii) Following DNI [68], we interpolate the weights between the identical model and the fine-tuned model.”
- Deep Learning for Image Super-Resolution: A Survey (2019, unknown): “[103], [155] propose a network interpolation strategy.”
- Exploiting Deep Generative Prior for Versatile Image Restoration and Manipulation (2020, methodology): “Inspired by [38], to morph between xA and xB , we apply linear interpolation to both the latent”