Arbitrary Upscaling

Arbitrary upscaling aims to increase the resolution of data, such as images or geological models, to a desired scale without being limited to fixed scaling factors. Current research heavily utilizes deep learning, particularly convolutional neural networks, often employing joint optimization strategies for simultaneous upscaling and downscaling to improve accuracy and robustness. This capability is crucial for diverse applications, including improving the comparability of remotely sensed data with ground measurements, enhancing the efficiency of geological simulations, and achieving higher-quality image rescaling in various fields.