<p>Outdoor weather conditions affects the visual quality of the high-definition image in surveillance and autonomous navigation applications. Rain streaks create the diverse appearance and blurring of the object in outdoor captured image or video sequences. So, single-image deraining is a crucial preprocessing step in many high-level computer vision applications to enhance the visual quality of real-time images. The state-of-the-art methods attempt to remove the non-statistical rain patterns and achieve moderate computational cost. Moreover, the relationship between background features and rain patterns is usually ignored during network training. Recently, high-performance self-attention mechanism has been used in most of the convolutional neural networks to exhibit superior restoration performance. But, self-attention framework suffers from computational cost, high inference time, and instability to capture the fine details during the network training. In response to these challenges, we propose a deep, fast-track single-image deraining framework to solve the above issues using an efficient spatial information attention module with generative adversarial network. We adopt basic spatial information attention module (M1) and directional spatial attention module (M2) to handle the rain pattern in different directions. Both (M1 and M2) modules confirm that extracting the attractive feature at a different scale improves the restoration process. Additionally, we introduce a specialized generator and pyramidal discriminator in our fast-track deraining network for better performance results. The enhanced generator network consists of the feature capture module and ground-truth attention module to learn network mapping between rainy and derained images. The average PSNR value of the proposed deep deraining network improves from 2.5 to 4.5% compared to other state-of-the-art methods. Experimental results demonstrates that our proposed work outperforms other existing state-of-the-art methods.</p>