Please add further explanations on the limitation of existing CNN-based reconstruction methods to highlight the motivation of the work.

Existing CNN-based reconstruction methods have proven effective in producing high-quality images from low-quality inputs. However, there are several limitations to these methods that motivated the need for further research.

Firstly, existing CNN-based reconstruction methods are often limited by the size of the training dataset. With limited training data, the resulting reconstruction may not generalize well to unseen data, leading to poor performance on new inputs.

Secondly, these methods may struggle to reconstruct complex structures and details due to the limited receptive field of the CNN architecture. This can result in blurry or distorted images, particularly when dealing with small or intricate features.

Thirdly, CNN-based reconstruction methods are often computationally expensive and may require significant resources to run in real-time. This can limit their practical application in real-world scenarios.

Lastly, these methods may be sensitive to noise and artifacts in the input data, which can cause the resulting reconstruction to be inaccurate or incomplete.

Overall, the limitations of existing CNN-based reconstruction methods highlight the need for further research into more robust and efficient methods that can overcome these challenges and produce high-quality reconstructions even in challenging scenarios.