This repo reproduces the results in the paper "Single Image Interpolation Exploiting Semi-local Similarity", a ICASSP paper published in 2019 with state-of-the-art performance in model-based single image interpolation.
To read the full text of this paper, use this link.
For quickly understand this paper, you can read the slides at this link.
To reproduce the results on Table 1, simply run main_2x.m and main_3x.m file in Matlab。It supports parallel computing by default.
If you find this paper or the code helpful to your research, please cite:
@inproceedings{yu2019single,
title={Single image interpolation exploiting semi-local similarity},
author={Yu, Lantao and Orchard, Michael T},
booktitle={ICASSP 2019-2019 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)},
pages={1722--1726},
year={2019},
organization={IEEE}
}
and
@inproceedings{yu2019spatially,
title={When Spatially-Variant Filtering Meets Low-Rank Regularization: Exploiting Non-Local Similarity for Single Image Interpolation},
author={Yu, Lantao and Orchard, Michael T},
booktitle={2019 IEEE International Conference on Image Processing (ICIP)},
pages={200--204},
year={2019},
organization={IEEE}
}