Abstract: Existing convolutional neural network-based reversible data hiding (RDH) predictors typically stack the standard convolution blocks with stride 1 for feature extraction, and keep the sizes ...
Abstract: Single image dehazing is a challenging ill-posed problem which estimates latent haze-free images from observed hazy images. Some existing deep learning based methods are devoted to improving ...
Some results have been hidden because they may be inaccessible to you
Show inaccessible results
Feedback