Learning to Explain: a Gradient-based Attribution Method for Interpreting Super-Resolution Networks.
Anni YuYu-Bin YangPublished in: ICASSP (2023)
Keyphrases
- super resolution
- high resolution
- low resolution images
- high resolution images
- low resolution
- high efficiency
- single image super resolution
- image super resolution
- super resolution reconstruction
- three dimensional
- motion estimation
- unsupervised learning
- computationally efficient
- test images
- single frame
- method can effectively reduce