Abstract: Although the vision transformer-based methods (ViTs) exhibit an excellent performance than convolutional neural networks (CNNs) for image recognition tasks, their pixel-level semantic ...
Abstract: Single image defogging that aims to restore a fog-free image from its appropriately unconstrained hazy environment is a fundamental yet complex work that has recently achieved enormous ...
Abstract: Haze obscures remote sensing images, hindering valuable information extraction. To this end, we propose RSHazeNet, an encoder-minimal and decoder-minimal framework for efficient remote ...
Some results have been hidden because they may be inaccessible to you
Show inaccessible results