SAM

We proposed a stereo attention module (SAM) to extend arbitrary single image SR methods for stereo image SR. [PDF]

We release the demo of SRResNet+SAM in a new branch. [Code]

Architecture of SAM

Visualization of Generated Attention Maps

Quantitative Results

Qualitative Results

Citiation

If you find our work helpful, please cite:

@article{SAM,
  title={A Stereo Attention Module for Stereo Image Super-Resolution},
  author={Ying, Xinyi and Wang, Yingqian and Wang, Longguang and Sheng, Weidong and An, Wei and Guo, Yulan},
  journal={IEEE Signal Processing Letters},
  year={2020},
  volume={27},
  pages={496-500},
  publisher={IEEE}
}

Related Work

Non-Local Nested Residual Attention Network for Stereo Image Super-Resolution, ICASSP 2020, [pdf]
Stereoscopic Image Super-Resolution with Stereo Consistent Feature, AAAI 2020. [pdf]
Flickr1024: A Large-Scale Dateset for Stereo Image Super-Resolution, ICCVW 2019. [pdf], [website].
Learning Parallax Attention for Stereo Image Super-resolution, CVPR 2019. [pdf], [code].
Enhancing the spatial resolution of stereo images using a parallax prior, CVPR 2018. [pdf].

Contact

You can contact us at yingxinyi18@nudt.edu.cn or wangyingqian16@nudt.edu.cn for any question about this work.

lz-xinlin/SAM

SAM