Self-Tuned Deep Super Resolution

Zhangyang Wang, Yingzhen Yang, Zhaowen Wang, Shiyu Chang, Wei Han, Jianchao Yang, Thomas S. Huang

CVPRW 2015 pp. 1-8

doi:10.1109/CVPRW.2015.7301266 /cvprw/2015/wang2015cvprw-selftuned/

Abstract

Deep learning has been successfully applied to image super resolution (SR). In this paper, we propose a deep joint super resolution (DJSR) model to exploit both external and self similarities for SR. A Stacked Denoising Convolutional Auto Encoder (SDCAE) is first pre-trained on external examples with proper data augmentations. It is then fine-tuned with multi-scale self examples from each input, where the reliability of self examples is explicitly taken into account. We also enhance the model performance by sub-model training and selection. The DJSR model is extensively evaluated and compared with state-of-the-arts, and show noticeable performance improvements both quantitatively and perceptually on a wide range of images.

PDF CVPRW Semantic Scholar

Cite

Text

Wang et al. "Self-Tuned Deep Super Resolution." IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops, 2015. doi:10.1109/CVPRW.2015.7301266

Markdown

[Wang et al. "Self-Tuned Deep Super Resolution." IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops, 2015.](https://mlanthology.org/cvprw/2015/wang2015cvprw-selftuned/) doi:10.1109/CVPRW.2015.7301266

BibTeX

@inproceedings{wang2015cvprw-selftuned,
  title     = {{Self-Tuned Deep Super Resolution}},
  author    = {Wang, Zhangyang and Yang, Yingzhen and Wang, Zhaowen and Chang, Shiyu and Han, Wei and Yang, Jianchao and Huang, Thomas S.},
  booktitle = {IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops},
  year      = {2015},
  pages     = {1-8},
  doi       = {10.1109/CVPRW.2015.7301266},
  url       = {https://mlanthology.org/cvprw/2015/wang2015cvprw-selftuned/}
}