Computer Vision algorithms easily get affected by images or videos captured in outdoor environments due to various bad weather conditions such as rain, fog, snow, haze. Two-stage deep neural networks based on attention learning are proposed for single image rain removal. Based on the fact that the implicit connection among rain lines within the image is higher than that between the rain line and the background image, it is easier to learn the rain lines in the rainy image using attention learning. The proposed Two-stage Attention-based Deep Neural Network (TA-DNN) for single image rain removal essentially consists of modules such as Inception, Sequential Dual Attention Block (SDAB), and Multi-Scale Feature Aggregation Module (MSFAM) for feature extraction, rain line detection, and transfiguration, respectively. The experimental results illustrate that the proposed method performs better when compared with the state-of-the-art methods both qualitatively and quantitatively.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.