RIRGAN: An end-to-end lightweight multi-task learning method for brain MRI super-resolution and denoising.
Miao YuMiaomiao GuoShuai ZhangYuefu ZhanMingkang ZhaoThomas LukasiewiczZhenghua XuPublished in: Comput. Biol. Medicine (2023)
Keyphrases
- lightweight
- end to end
- super resolution
- multi task learning
- high resolution
- low resolution images
- brain mri
- denoising
- low resolution
- segmentation method
- theoretical analysis
- computer vision
- image reconstruction
- motion estimation
- machine learning
- expectation maximization
- image quality
- image data
- probabilistic model
- closed form
- pairwise
- prior information
- noisy images
- computational complexity
- high quality