DEA-Net: Single Image Dehazing Based on Detail-Enhanced Convolution and Content-Guided Attention

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 the model performance via increasing the depth or width of convolution. The learning ability of Convolutional... View More