Enhancing images captured in in reduced illumination is a difficult problem within computer vision, primarily due to reduced visibility, color distortion, and noise amplification during enhancement. This paper presents LEARN, a novel Laplacian Enhanced Attention and Residual Network architecture integrated with the Convolutional Block Attention Module (CBAM), offering improved feature refinement through spatial and channel attention mechanisms for effective low-light image enhancement. Our approach integrates a Laplacian enhancement module that preserves and enhances edge details, expanded kernel residual blocks that increase the receptive field for better contextual information, and CBAM attention mechanisms that focus on relevant features while suppressing noise. Extensive experiments demonstrate that LEARN achieves competitive performance, reaching an average PSNR of 23.34 dB and SSIM of 0.8613 across LOLv1 and LOLv2 real and LOLv2 synthetic datasets. Our model maintains a relatively lightweight architecture while effectively enhancing image brightness, preserving structural details, and maintaining natural colors. Both quantitative and qualitative findings verify LEARN's effectiveness in tackling the difficulties associated with enhancing images captured under diverse lighting conditions.